Appendix
Volume I
The
Belmont
Report
Ethical Principles
and Guidelines for
the Protection of
Human Subjects
of Research
The National Commission
for the Protection of Human Subjects
of Biomedical and Behavioral
Research
gatjsesda, Msrytmi 2B6U
Appendix
Volume I
The
Belmont
TReport
Ethical Principles
and Guidelines for
the Protection of
Human Subjects
of Research
The National Commission
for the Protection of Human Subjects
of Biomedical and Behavioral
Research
This Appendix contains (in two volumes)
the full text of the papers that were prepared
to assist the Commission in its consideration
of the basic ethical principles that should
underlie the conduct of research
involving human subjects.
DHEW Publication No. (OS) 78-0013
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402
TABLE OF CONTENTS
APPENDIX TO BELMONT REPORT
Volumes I and II
Volume I
I. PRELIMINARY PAPERS PREPARED FOR THE COMMISSION
BY ROBERT J, LEVINE, M.D.
1. The Boundaries Between Biomedical or
Behavioral Research and the Accepted
and Routine Practice of Medicine
2. The Role of Assessment of Risk Benefit
Criteria in the Determination of the
Appropriateness of Research Involving
Human Subjects
3. The Nature and Definition of Informed
Consent in Various Research Settings
4. Appropriate Guidelines for the Selection
of Human Subjects for Participation in
Biomedical and Behavioral Research
II. BASIC ETHICAL PRINCIPLES RELATING TO RESEARCH
INVOLVING HUMAN SUBJECTS
5. Ethical Principles and Their Validity Kurt Baier, D. Phil.
6. Distributive Justice and Morally Relevant
Differences Tom Beauchamp, Ph.D.
7. The Identification of Ethical Principles James Childress, B.D., Ph.D.
8. Basic Ethical Principles in the Conduct of
Biomedical and Behavioral Research
Involving Human Subjects H. Tristram Engelhardt,
Jr., Ph.D., M.D.
9. Medical Ethics and the Architecture of
Clinical Research Alvan R. Feinstein, M.D.
Jeffrey L. Lichtenstein,
M.D.
10. How to Identify Ethical Principles Alasdair Maclntyre, M.A.
Belmont Appendix
11. Some Ethical Issues in Research Involving
Human Subjects
LeRoy Walters, B.D.
Ph.D.
Volume II
III. BOUNDARIES BETWEEN RESEARCH AND PRACTICE
12. Protection of the Rights and Interests
of Human Subjects in the Areas of Pro-
gram Evaluation, Social Experimenta-
tion, Social Indicators, Survey Re-
search, Secondary Analysis of Research
Data, and Statistical Analysis of Data
From Administrative Records . . Donald T. Campbell, Ph.D.
Joe Shelby Cecil , Ph.D.
13. Response to Commission Duties as Detailed
in P.L. 93-348, Sec. 202(a) (1 ) (B) (i ) . . . Donald Gallant, M.D.
14. On the Usefulness of Intent for Distinguishing
Between Research and Practice, and Its Replace-
ment by Social Contingency Israel Goldiamond, Ph.D,
15. Boundaries Between Research and Therapy,
Especially in Mental Health Perry London, Ph.D.
Gerald Klerman, M.D.
16.
17.
Legal Implications of the Boundaries
Between Biomedical Research Involving
Human Subjects and the Accepted or
Routine Practice of Medicine. ....
John Robertson, J.D.
The Boundaries Between Biomedical Re-
search Involving Human Subjects and
the Accepted or Routine Practice of
Medicine, with Particular Emphasis on
Innovation in the Practice of Surgery
What Problems are Raised When the Current
DHEW Regulation on Protection of Human
Subjects is Applied to Social Science
Research?
David Sabiston, M.D.
Richard A. Tropp
Belmont Appendix
IV, RISK/BENEFIT CRITERIA
19. Some Perspectives on the Role of Assess-
ment of Risk/Benefit Criteria in the
Determination of the Appropriateness of
Research Involving Human Subjects , , , Bernard Barber, Ph.D.
20. The Role of Risk/Benefit Analysis in the
Conduct of Psychological Research Gregory Kimble, Ph.D.
21. A Philosophical Perspective on the Assess-
ment of Risk-Benefit Criteria in Connection
with Research Involving Human Subjects Maurice Natanson, Ph.D.
22. Essay on Some Problems of Risk-Benefit
Analysis in Clinical Pharmacology Lawrence C. Raisz, M.D.
INFORMED CONSENT
23. Nature and Definition of Informed
Consent in Research Involving Deception ..... Diana Baumrind, Ph.D.
24. Some Complexities and Uncertainties
Regarding the Ethical ity of Deception
in Research with Human Subjects Leonard Berkowitz, Ph.D.
25. Selected Issues in Informed Consent and
Confidentiality with Special Reference
to Behavioral/Social Science Research/
Inquiry Albert Reiss, Jr., Ph.D.
26. Three Theories of Informed Consent:
Philosophical Foundations and Policy
Implications , Robert Veatch, Ph.D.
I
PRELIMINARY PAPERS PREPARED FOR THE COMMISSION
BY ROBERT J. LEVINE, M.D.
1
THE BOUNDARIES BETWEEN BIOMEDICAL OR BEHAVIORAL
RESEARCH AND THE ACCEPTED AND ROUTINE
PRACTICE OF MEDICINE
Robert J. Levine, M.D.
July 14, 1975
The Commission is charged with the responsibility to consider, among
other things, (i) The boundaries between biomedical or behavioral research
involving human subjects and the accepted and routine practice of medicine.
It is fortunate that sharp definitions of the boundaries are not required.
Even a superficial exploration of this problem (contained in this paper) will
reveal the impossibility of describing mutually exclusive subsets (one called
research and one called practice) of the universe of activities in which health
care professionals may be engaged. It will be possible to describe some
activities as pure research and some other activities as pure practice; these
activities will be defined but not discussed in this paper. The focus of this
paper will be on activities having one or more of the following characteristics:
1) there may be legitimate dispute as to whether they are research or practice;
2) there is no dispute that they are combinations of research and practice; or
3) there may be some confusion in the view of one or more of the participants
in the process as to whether the activity is either research or practice or
a combination thereof.
The purpose of the considerations
What is the point in considering the boundaries between research and the
practice of medicine? To what use will such considerations be put? Let us
first agree that we do not know the answers to these questions. Alternatively,
let us first agree that each party to these considerations is likely at the
outset to answer these questions somewhat differently from each other. There
are those who assume that the final recommendations of the Commission will in-
dicate that different sorts of protections are required for subjects (the sub-
jects of the research activities of investigators) than are required for patients
1-1
(the clients of health care professionals). To that others will respond that
the Commission has been further charged to "...consider the appropriateness of
applying the principles and guidelines identified and developed under subpara-
graph (A) to the delivery of health services to patients...". (This issue will
be the subject of another staff paper.)
There are those who will assume that the requirements for informed con-
sent are greater for subjects than they are for patients. To this one may
respond that in many recent malpractice litigations the physician has been
found "negligent" for having failed to provide "full disclosure". All of these
cases have been in the context of practice, not research.
Another assumption from which one might proceed is that it is necessary
to identify research so that a determination might be made as to what sorts of
activities should be reviewed by institutional review boards. Technically, this
assumption cannot now be challenged. Yet there are in most institutions devoted
to practice various panels, committees, and boards similarly charged to review
the practice activities within those institutions. Ongoing review of practice
activities is beginning under Professional Standards Review Organizations (PSROs).
So, perhaps one purpose of these considerations is to determine which activities
are the proper turf of which review body.
There are many other assumptions that might be stated and challenged at
the outset. They all have in common that research could or should be seen as
somehow differing from practice. As the Commission focuses on its primary purpose-
that contained in its title--the Protection of Human Subjects — the ultimate focus
of this paper becomes more clear. The purpose of defining the boundaries between
research and practice is to determine where one might find research subjects to
1-2
protect. The extent to which the protections required for research subjects
may differ from those required for patients remains to be determined. This
paper is written with the assumption that patients and subjects should be
treated somewhat differently. At the very least the individual should be
aware of which role he (or she) is playing--or being asked to play--and what
the implications are of playing the role of subject as opposed to patient.
Patients and subjects ( definitions )
At this point the words patient and subject--as they are used in this
paper--will be defined; then there will be a brief statement of some of the
assumptions that derive from these definitions. As the paper develops the
distinctions between patients and subjects should become increasingly clear
and the ramifications of the assumptions contained in these definitions should
seem increasingly important.
A patient is a client of a health care professional. By contrast, a
subject is an individual who is observed or experimented with by an investigator.
In determining the interactions between the individual (patient or subject) and
the professional (health care professional or investigator) the final decision-
making authority resides in the non professional. However, the patient may
ordinarily assume that the health care professional will be acting in his (or
her) best interests. Thus, in ordinary circumstances, the patient may choose
to delegate decision-making authority to the health care professional to what-
ever extent he (or she) wishes. Ideally, he should be aware that he may withdraw
this delegation of authority whenever it seems appropriate. For example, he
1-3
might seek consultation with another health care professional. The responsible
health care professional will, in fact, advise the patient when it might be
appropriate to withdraw delegation of decision-making authority for any of a
variety of reasons. The subject, on the other hand, should feel very less secure
in delegating decision-making authority to the investigator. While the health
care professional might be assumed to see the well-being of the patient as the
most important end, the investigator is assumed to see development of new
knowledge as a major, if not the ultimate, end. Thus, there is the ever-present
possibility that the investigator may see the development of new knowledge as an
end that takes precedence over the well-being of the subject. To the extent that
this happens, the role of the subject increasingly approximates that of a means
to another end. The subject is always at least in part a means and, in extreme
cases, is entirely a means and not an end. Thus, the subject can almost never
confidently delegate decision-making authority to an investigator.
Boundarie s of the considerations
This paper will begin with a discussion based upon a literal interpretation
of the minimum charge to the Commission. That is, there will first be a con-
sideration of the boundary problems that arise in the course of the activities
of those individuals who are licensed to conduct the "accepted and routine practice
of medicine". These individuals are physicians (including psychiatrists); an ex-
ploration of their activities will reveal most of the boundaries between both
biomedical and behavioral research on one hand and the accepted and routine
practice of medicine on the other. However, it is acknowledged at the outset,
that there are many other types of health care professionals; their activities
1-4
will be discussed later. The concentration on the activities of physicians
at the outset may, to a large extent, be considered paradigmatic.
There will also be a discussion of boundary problems that may arise in
manipulations of the health delivery system itself.
Other topics that will be covered include the following: There will be
a discussion of some of the special sorts of problems that arise when research
is conducted within the health delivery system by individuals who are not
health care professionals. Further, there will be a discussion of the boundary
problems created by introduction of new sorts of health "professionals" into
the health delivery system. There will be a discussion of research that has
no legitimate recognizable boundary with the accepted and routine practice of
medicine. Specifically, this discussion will include some behavioral research
and most social science. It will be demonstrated that the conceptual models
used to distinguish research from practice in the medical realm may be applied
to a limited extent to considerations of boundaries between the research and
"practice" activities of social scientists. Finally, there will be a statement
with little discussion that the conceptual models used to distinguish the boundaries
between research and practice in biomedical and behavioral research might be
applied fruitfully to considerations of analagous activities conducted in education
and in welfare.
Conceptual models
There are two traditional models for distinguishing research from practice.
The tradition of the professional (physician-investigator)--particularly the
professional acting as an individual—is to distinguish research from practice
on the basis of intent (of the professional). The second traditional system of
classification is that generally practiced by groups (ranging from Peer Groups to
1-5
Regulatory Agencies). Groups make the distinction on the basis of acceptance
or approval (by the group). It should be understood at the outset that — al-
though these two systems of classification constantly interplay with each other—
for purposes of discussion it is convenient to attempt to dissect them apart.
There will first be an exploration of the utility of distinguishing research
from practice on the basis of intent. This will be followed by a discussion of
the devices of acceptance and approval. An attempt will be made to explore the
tensions and conflicts created by the coexistence of these two models.
Boundaries defined by intent
Using intent as the taxonomic device, the activities of physicians may be
defined as follows. If a physician proceeds in his interaction with a patient
to bring what he considers to be the best available technique and technology to
bear on the problems of that patient with the intent of doing the most possible
good for that patient, this may be considered the pure practice of medicine. By
contrast, if a physician interacts with an individual with the intent of de-
veloping new knowledge (not primarily for the benefit of that individual), this
activity may be classified as research. The parenthetical clause in the preceeding
statement is meant to accommodate the understanding that diagnostic testing— which
may or may not be done with research intent— develops new knowledge; if the pur-
pose of the new knowledge is to benefit the individual directly, the activity is
classified as practice, not research. In a broader context research (involving
humans) may be defined as follows:
Research (involving humans) is any manipulation, observation, or other
study of a human being- -or of anything related to that human being that might
1-6
subsequently result in manipulation of that human being--done with the intent
of developing new knowledge and which differs in any way from customary medical
(or other professional) practice. Research need not be interactive; eg, ob-
servations of humans through a one-way glass, by tape-recording their conversations
with each other, or by examining their records may--but need not necessarily--
be classified as research.
Research may usually be identified by virtue of the fact that is conducted
according to a plan. The plan (protocol) will usually indicate the intent
(purpose) of the planner in conducting the activity. However, even if a plan
does not exist, activities that might be considered research (as defined in the
preceeding paragraphs): 1) should be considered research; 2) ordinarily should
be planned; and 3) should be reviewed by an IRB.
The word plan as used in the preceeding paragraph means a formal plan con-
taining a stated hypothesis, methods of procedure, and so on. It is acknowledged
that almost all meaningful activity is conducted according to some sort of plan
be it formal or informal. In many kinds of research—particularly in its early
stages — some sort of plan may exist in the mind of the investigator who may or
may not be able to state it formally. In biomedical research — particularly that
involved in elucidating the "natural history" of various diseases— the initial
impetus to the development of a plan most commonly is serendipitous. After an
initial observation is made, the physician-investigator attempts to make similar
observations on similar patients. A second common approach to developing the
first findings bearing on the natural history of a disease is as follows : An
investigator may develop a new test for say a new enzyme or other substance in
some body fluid. He may then proceed to measure this substance rather indis-
1-7
criminantly--if the intervention is relatively harmless—in the entire dis-
eased (and sometimes normal) population available to him. Thus, when one
investigator learned how to assay serotonin in body fluids he proceeded to
do so in virtually every patient in the institution in which he was working.
One patient had very high levels--and so on. Thus, the biochemical basis
for the malignant carcinoid syndrome was discovered. As with most such
research, when the final report appeared, it seemed quite rational to have
deliberately set out to look for this biochemical abnormality in this disease.
The research plan in this second case--owing to its similarities to Darwin's
experiences- -is commonly called a Voyage of the Beagle.
Similarly, social scientists may or may not begin with clearly stated
hypotheses or easily articulated plans. Thus, for example, although anthro-
poligists and ethnologists have a clear understanding of their research
strategies, they ordinarily cannot state their hypotheses or provide detailed
accounts of how they will apply their general strategies to their particular
projects until they have begun. Similarly, even when a research plan is
formally and clearly articulated at the outset by a sociologist, it may
evolve dramatically and rapidly after the first few "experiments" or interviews
are completed. In the early stages of social research the plan may develop
heuristically as unexpected and "even more interesting" possibilities are
discovered in the research environment.
Complex and simple activities
For purposes of defining problem areas, activities will be categorized
as either complex or simple. A complex activity is one in which the interaction
between the physician and the individual (patient or subject) involves more
than one maneuver some of which may be done with research intent and others
with practice intent. A simple activity may be defined as a single maneuver
the intent of which may be either research or practice.
Complex activities
First there will be a consideration of some examples of boundary problems
that may arise in the conduct of complex activities. In general, the boundary
problem here entails sorting out which parts of the activities are research
and which are practice. Thus, if a complex activity is divided up into its
component parts, one may find that of 100 simple activities only 2 or 3 are
research. The reason to identify these is to enable the professional to
inform the patient- subject as to which are which so that he might choose (for
his own reasons) to reject some with a full awareness of the potential con-
sequences of such rejection or acceptance. The discussion of complex activities
which follows is intended to provide a basis for discussing models of informed
consent in a subsequent staff paper.
1) In dealing with a patient with an unusual disease a physician may
perform more diagnostic tests than are necessary to provide optimal care. The
purpose of the excessive diagnostic testing might be to develop a very com-
prehensive description of the individual with an ultimate goal of publishing
a "case report". In this situation the individual has usually contacted the
physician with the expectation that he will be playing the role of patient.
However, full disclosure would demand that the physician inform the individual
that he is being asked to also play the role of subject. The individual should
be informed that some of the proposed diagnostic tests are to be done with re-
1-9
search intent. Further, he should be informed that these tests are not necessary
to provide optimal care; that refusal to accept these tests will in no way ad-
versely prejudice the ability of the physician to provide optimal care; and that
(ordinarily) the subject is not expected to pay for these tests.
2) A frequently encountered boundary problem exists when research is de-
signed to compare the safety and/or efficacy of two "accepted" therapeutic or
diagnostic modalities. While the use of either of these modalities in the con-
text of a physician-patient relationship might not be considered experimental,
a study designed to compare them is. Thus, full disclosure demands that the in-
dividual be informed that he is being invited to play (at least in part) the role
of subject and what implications this may have. Among other things, during the
course of the studies the subject will be subjected to more testing and closer
scrutiny than would be the case if the same maneuvers were implemented as part
of the usual practice of medicine. Further, the prospective subject should be
informed that some practitioners of medicine may have strong feelings that one
of the maneuvers being studied is clearly superior to the others and what the
implications are of these strong feelings. Thus, for example, a subject may
encounter in a social environment (eg, cocktail party) a physician who might
express such strong feelings : What sort of lunatic talked you into taking
drug B for your condition when everybody knows that drug A is superior?
Simple activities
Let us next consider some of the boundary problems that arise in con-
siderations of simple activities. Here one is most concerned with the difficulties
that might arise in providing satisfactory definitions of what constitutes
1-10
"accepted" or "best available" technique and technology. Here one is parti-
cularly concerned that the physician may use a technique convinced that it is
"accepted" and "best available" while an appropriately constructed panel of
peers or a regulatory agency might not agree. Some of the boundary problems
in this area will be exposed through consideration of the use of drugs in the
practice of medicine. It will also demonstrate that our definition of practice
(as distinguished from research) contains within it a boundary problem; that is,
when the concept of intent comes into conflict with the concept of what con-
stitutes accepted or best available technology.
Conflicts between intent and acceptance or approval
If a physician wishes to use a drug that does not yet have a New Drug
Application (NDA) approved by the Food and Drug Administration (FDA) this
is--in the view of the FDA--research no matter what the intent of the physician.
The same situation obtains if a drug with an approved FDA is used for an unapproved
purpose. There are at least two problems subsidiary to this which should be
studied in some detail:
The first is the conflict between the existing and proposed DHEW rules
and policies governing experimentation on children and pregnant women on one
hand, and, on the other, the FDA requirements for proof that a drug is safe and
effective within these populations. As a consequence of this conflict in re-
gulations and policies, the vast majority of drugs contain warnings on their
FDA-approved package labels that "the safe use of this drug in children or in
pregnant women has not been established", or other similar caveats. Thus, one
might consider most drug use in these population groups to be either uncontrolled
(unethical ?) research or (potentially) illegal practice of medicine.
The second is that in the "experience" of the practicing community a drug
I'll
may become identified as the "drug of choice" in a specific situation long
before this use is approved by the FDA. In some cases this "experience" is
based upon anecdotal reports or hearsay evidence. In others, there may develop
a substantial body of evidence appropriately accumulated, reported, and debated
which seems--in the view of physicians expert in the field--to support the
identification of the agent as the "drug of choice". Examples include the use
of lidocaine in the treatment of some arrhythmias; the use of propranalol and
diazoxide in the treatment of hypertension, and so on. FDA recognition of
such uses for such drugs may lag many years behind their acceptance in the
community of practitioners.
Thus, for purposes of considering drugs as either accepted or not accepted
there is the FDA which helps solve some boundary problems and which may contribute
to the creation of some others. There are a variety of other types of therapeutic
and diagnostic modalities for which we have no standard- setting agencies. For
these, for the time being, one might consider developing mechanisms to devise
guidelines which will help practitioners distinguish accepted from unaccepted.
By implication, it may very well be that the best mechanism would not entail
development of new federal regulatory agencies. Among the areas which might be
assigned high priority at this point are the following: surgery, invasive dia-
gnostic techniques (including biopsies and catheterizations), diagnostic and
therapeutic devices, radioisotopes, radiation and other physical forms of therapy,
and behavioral approaches to psychotherapy.
The FDA model
Another sort of problem that arises in considering the status of simple
activities is that of determining when an activity passes from the realm of
research into that of "customary or best available" or "approved or accepted"
1-12
technique. Here it may be useful to explore the approach used by the FDA
in relation to drugs to see how it might be applied to the other categories
of simple activities mentioned in the preceeding paragraph. The development
of a new drug proceeds along a spectrum which contains several discontinuities.
Thus, in the development of a drug, the initial testing in humans is called
"Phase I". Standards are set that permit filing of a INDA which, in turn,
permits moving from the pre-clinical to clinical experimentation stage.
Standards are further set to determine when one may proceed from Phase I to
Phase II and then to Phase III. Procedures are available to allow simultaneous
conduct of these phases when appropriate. Perhaps it would be appropriate to
see how this sort of phasing might be used in the development of, for example,
a surgical technique or a therapeutic device. In consideration of boundaries
the point of discontinuity is the passage from Phase III to approval of the
drug for general distribution. This is signified by filing with and approval
by the FDA of a NDA. One should be concerned with both sides of the boundary.
Thus, when is it dangerous (or unethical) to consider a simple activity "standard
or accepted" when it has not yet been suitably tested? Alternatively, when is it
inappropriate (or perhaps dangerous) to continue research on a simple activity
that has already been "proved"? In the latter case, is it appropriate to deprive
individuals of this technique by further testing in "controlled" studies? It
should be noted that the FDA has come to terms in a responsible manner with the
understanding that there are, at times, drugs which clearly should be made
available to individuals with serious illnesses but which have not yet passed
full criteria for issuing a NDA. Under these circumstances a limited NDA may
be issued through which only physicians who are acce-ptsble to the FDA may prescribe
1-13
the drug and under which they assume the obligation to continue to collect
data on safety and efficacy in conjunction with its use.
Innovative therapy
At this point it may be useful to examine the concept of innovative
therapy. Innovative therapy is a term applied to a simple activity that is
ordinarily conducted by the physician with either pure practice intent or
varying degrees of mixed research and practice intent. It is distinguished
from accepted and standard medical practice in that it has not been sufficiently
tested to meet peer group or regulatory agency standards for acceptance or
approval. To return to the drug (FDA) model, innovative therapy would be those
activities classified as phase II or phase III drug trials. Ordinarily, such
drug trials are done with mixed intent. There is the intent of alleviating or,
perhaps, curing a disease. At the same time, there is the intent to develop
new knowledge in a systematic way to determine whether this drug is better
(more safe or more effective) than other drugs that might be available for the
same purpose. After a sufficient number of such interventions have been per-
formed the drug may be classified as accepted in which case the word, innovative,
is no longer necessary in describing its use as therapy.
In other cases innovative therapy may be conducted with pure practice in-
tent. Thus a physician may decide to administer a drug to a patient who has a
serious abnormality requiring treatment for which there is no alternative. It
might be that there is no other drug or other form of treatment available for
this condition. Alternatively, it might be the case that alternative therapeutic
modalities have been tried and failed. Thus, the physician may proceed with pure
practice intent. In some cases the physician might not perceive himself as an
1-14
investigator and might be irritated by the fact that he cannot have access to the
drug without completing certain forms designed to develop in a systematic fashion
the new information required to document the relative safety and/or efficacy of
this drug.
The term, innovative therapy, applies to much more than drug administration.
However, it is in the area of drug administration that there now exists the
most highly developed regulation. Other categories of innovative therapy in-
clude surgery, invasive diagnostic techniques (including biopsies and cathe-
terizations), diagnostic and therapeutic devices, and so on as stated earlier.
Timely removal from the market
From the opposite point of view we should be concerned with mechanisms
for formally removing an activity from the category of "standard and accepted".
Again, it might be appropriate to study FDA procedures in this regard. Thus
we might learn of procedures they have available for surveillance of ongoing
activities which allow them to discover when certain solutions for intravenous
administration should be removed from the market because they are contaminated
with bacteria (temporary removal) or when a drug should be partially removed
from the market (have its market limited) owing to the discovery of compli-
cations not understood at the time it was approved (eg, tolbutamide and chlor-
amphenicol). Thus, we might discover devices for timely removal from the
"market" of some surgical procedures (eg, internal mammary artery implantations
and thymectomies "for convenience"), use of fluoroscopy for fitting shoes,
use of routine catheterizations of the urinary
bladder for diagnostic purposes, and so on.
Caution about extending the FDA model
In many cases the safety and efficacy of an intervention is directly re-
1-15
lated to the personal skill and experience of the individual who is performing
it. This is most conspicuously true in some sorts of surgery. Thus, for ex-
ample, when a highly skilled and experienced team of cardiovascular surgeons
report the results of their experience with a specific operation one cannot
assume that the same safety and efficacy will be achieved by less skilled
and/or experienced surgeons. This is at least equally- -albeit less con-
spicuously—true in considerations of behavioral approaches to psychotherapy.
To some extent the personal skills and experience of the physician will in-
fluence the outcome of his use of any diagnostic or therapeutic modalities.
Parenthetically, it should be noted that personal skills and experience
greatly influence the outcome of the use of drugs. Thus, for example, the
use of methotrexate by an experienced oncologist may yield salutary effects
with minimal risk while its use by many other physicians may be deadly.
Another boundary issue that might be explored involves the psychological
and other intrapersonal conflicts that develop when one individual attempts
to play both roles--that of physician and that of investigator-- simultaneously.
We doubt that we can do much to resolve these conflicts. However, the final
report of the Commission should at least acknowledge its existence and its
implications. Among the individuals who have published their analyses of
these conflicts are W. Bean: J. Lab. Clin. Med. 39: 4, 1952. and J. Katz:
Clin. Research. 21: 787, 1973.
Other health care professionals
As mentioned earlier physicians are not the only health care professionals
who might be considered. The foregoing is a partial list of health care pro-
1-16
fessionals whose activities involve direct interface with individuals (patients
and/or subjects). The list is not comprehensive and is meant merely to illustrate
the diversity with which one might deal:
Osteopaths, naturopaths, chiropracters , dentist, podiatrists, chiropodists,
pharmacists, clinical psychologists, marriage counselors, psychoanalysts (non-
physician), counseling psychologists, social workers, dietitians, nurses (in-
cluding licensed practical nurses, registered nurses, nurse practitioners, nurse
clinicians, etc), physician's assistants (types A, B, and C) (NSF classification),
Medex, child health associates, orderlies, nurses aides, radiology technicians,
emergency medical technicians, and ambulance drivers.
For practical purposes these individuals can be classified in two categories :
1) Free-standing practitioners and 2) those who can act only by delegated authority
from an individual in the first category. Perhaps we should concentrate in our
considerations on the activities of individuals in category 1 with the assumption
that they can be held accountable for the activities of those individuals to whom
they delegate authority (respondeat superior). However, it should be understood
that there is considerable variation in the medical practice laws of various
states as to who may practice in category 1 and who may practice in category 2.
Currently, there is legislation pending in several states which would move some
individuals from category 2 to category 1. Further, it must be understood that
some individuals in category 2 may--in the context of practice--be called upon
to exercise varying degrees of independent judgment. It must also be understood
that most of the individuals in category 2 may have authority delegated to them
by a physician or another individual in category 1 who is engaged in research as
well as practice.
Thus, in considering the boundaries between research and practice the same
issues arise in considerations of these individuals as discussed earlier in
considerations of physicians. The specific activities th~. must be considered
1-17
are enlarged by adding to the list some sorts of individuals who engage in
health care activities in which physicians are not ordinarily involved; eg,
dentists, podiatrists, acupuncturists, etc.
New health professionals as experimental agents
Another problem one might consider is the fact that the introduction of
certain types of individuals into the health delivery system may itself be
considered experimental. This issue may be considered at several levels. The
first is the level at which a new type of health professional is created based
on the assumption that by virtue of a formal training experience this individual
will prove to be safe and effective in his prescribed function. It is difficult
to anticipate the future but, looking back just a few years, at which point might
one have considered the introduction of family nurse practitioners and physician's
assistants into health delivery systems experimental and at what point did we
begin to consider them part of the "accepted and routine practice of medicine".
Through reviewing these developments we might be able to formulate guidelines
for considering the present and future development of new types of allied health
personnel.
At another level one might consider the extent to which introduction of
new health personnel who have no formal training or certification might be
considered experimental. Here we are concerned with two types of problems.
The first is that a physician or other category 1 individual may undertake to
train any one he selects to perform part of the professional functions. Thus,
an orthopedic surgeon may train someone to apply casts, assign the title of
"cast-technician" and allow that individual to apply casts with or without direct
supervision. Similarly, other individuals may be trained to assume major health
1-18
care responsibilities; eg, diabetic teaching aides, genetics counselors, primal-
scream assistants, receptionist-history takers, and so on. A second problem
is introduction into a health delivery system of an individual who has no
training whatever based on the assumption that by virtue of his experience he
will be able to contribute constructively to the functions of the system. Thus,
for example, ex-drug addicts are introduced into drug dependency units (eg,
methadone maintenance programs), spouse-surrogates are introduced into sex
therapy teams, and so on.
Should we also be concerned with institutions (or individuals) that may
act outside the established health delivery system but which receive referrals
from within the system which are comprised of individuals of the sort described
as the second problem in the preceeding paragraph? For example, we might con-
sider organizations of people who have recovered from laryngectomies (Lost
Chord Clubs), colostomies (Q-T Clubs), Alcoholics Anonymous, and Parents
Without Partners.
Activities of investigators who are not health care professionals
The individual subject will not ordinarily be confused when approached by
an investigator who is not a health care professional (eg, a sociologist) as to
whether he is being asked to play the role of patient or subject. However, there
are circumstances in which the activities of investigators who are not health-
care personnel may create boundary problems. Two examples are offered for
purposes of illustration.
1) The first is the situation in which the investigator is conducting
his experiments in a health delivery setting. Thus, a sociologist interviewing
patients in a hospital may be perceived by the patient (subject pro tempore)
1-19
as an agent of the hospital. The patient may not understand that this part
of the total activities in the hospital is not part of the "practice of
medicine".
2) Researchers who are not health-care professionals may gain access
to prospective subjects through interactions with (eg, review of the records of)
a physician or a hospital. Thus, in this way a physiologist may contact indi-
viduals who have recovered from myocardial infarctions or a geneticist may
contact individuals with epilepsy. Again, the prospective subject may mis-
interpret the role of the investigators as perhaps being part of the health
care delivery team. Also, in this category major issues of confidentiality
are raised.
Manipulations of the system
To this point the discussion has focused on the activities of individuals
operating in health delivery systems. Next to be considered are the boundary
problems created by manipulations of the system itself. At what point is a
modification of the system considered experimental and at what point may it be
considered accepted and routine practice of medicine? A partial list of modi-
fications of the system follows: Health maintenance organizations (HMO), pre-
paid group practices, innovative social service programs, multiphasic screening,
computerized medical record systems, computerized approaches to diagnostic and
therapeutic decision making, experimental ambulance systems (e.g., coronary
care ambulances), and experimental communications networks. Just as with all
other activities discussed earlier, one might also consider here the rights of
the individual to know the extent to which he is a patient or a subject as well
as full disclosure of options, risks, benefits, and so on.
1-20
For example, fundamental to the concept of the HMO is that for economic
reasons (among others) more people will avail themselves of more health care
services. Does the provision of additional services increase the likelihood
of iatrogenic diseases? If so, does an ethical question arise? If an ethical
question arises does it follow that it might be resolved through regulation?
One might also consider the ethical and/or regulatory implications of
computerized diagnostic and therapeutic decision-making systems. The former —
based upon initial inputs—may proceed to order additional diagnostic testing
without "consulting" a physician. Similarly, the latter may proceed to
formulate and implement therapeutic decisions without "consulting" a physician.
Considerations of the boundaries of the charge to the Commission :
The preceeding pages have been concerned with explorations of the boundaries
between biomedical or behavioral research involving human subjects and the ac-
cepted and routine practice of medicine. This was based upon a tentative ac-
ceptance—as a matter of convenience— of a literal interpretation of the minimum
charge to the Commission as the extent of its obligation or prerogative to
explore. The assumption was discussed that some may believe that the human
subject of research requires more protection- -more stringent safeguards of
his rights and welfare- -than does the patient.
An equally tenable argument may be made that all human subjects of re-
search require more protection than do similar individuals playing alternative
roles that might be available to them. At the very least, they ought to be
made aware of the facts that they are being asked to play the role of research
subject, what risks and benefits might accrue to them from playing this role
1-21
and that ordinarily they have the right to refuse to be subjects without ad-
versely prejudicing their relationship to any portion of society that they
value. With that objective in mind, we may now consider briefly the boundaries
between the research activities and the "practices" of professionals in two
additional major categories: 1) Social sciences, and 2) education and welfare.
Social sciences :
Earlier in this paper there was a brief discussion of the activities of
social scientists as they were conducting their research either within or
in collaboration with the health delivery system. Those sorts of activities
of the social scientist present no boundary problem to the IRB. This is
clearly research. The boundary problem in that case is presented to the
patient. The patient may be confused about the role of the social scientist
and, derivitavely, about his own. Thus, unless the patient is informed that
he is being asked to assume the role of subject during the period of in-
teraction with the social scientist, he may consider these activities part
of the standard "practices" he is expected to participate in within the
health delivery system.
When social scientists are performing their activities outside the health
delivery system there is clearly no boundary problem between their activities
and the "accepted and routine practice of medicine".
Are there any boundary problems to be considered in the activities of
social scientists between their research and "practice" activities? First,
it seems appropriate to state that in considering the usual activities of
sociologists, anthropologists, political scientists, and so on, there is no
boundary problem since they have no professional practices other than research.
1-22
However, the activities of some social scientists will present us with sub-
stantial boundary problems between their research activities and professional
practices (it should be clear at this point that the scope of the word practice
now is extended deliberately beyond the definition contained in the charge to
the Commission). A major boundary problem might be presented by the activities
of a criminologist who also either is himself a law-enforcement officer or is
working in collaboration with a law-enforcement agency. Similar, though less
dramatic problems might be presented through considerations of the activities
of systems analysts, administrative scientists, and so on. Other sorts of
boundary problems are presented by the fact that some social scientists con-
tribute to the formation of public policy. As a consequence of their usual
professional activities (doing research) they may develop knowledge that will
be used to "remedy" the societal dysfunctions they have "diagnosed". Thus,
in this fashion, their role might be compared to that of the physician with
society as the patient- subject. Thus, perhaps in some cases, it is society
that we ought to offer the opportunity to give informed consent.
The definition of research based upon intent of the professional as
modified through considerations of "accepted" or "approved" provided earlier
in this paper might be equally applied to considerations of the activities
of social scientists: To reiterate:
Research (involving humans) is any manipulation, observation, or other
study of a human being- -or of anything related to that human being that might
subsequently result in manipulation of that human being--done with the intent
of developing new knowledge and which differs in any way from customary medical
1-23
(or other professional) practice. Research need not be interactive; eg, ob-
servations of humans through a one-way glass, by tape-recording their con-
versations with each other, or by examining their records may- -but need not
necessarily--be classified as research.
What one is most concerned with in the activities of the social scientist
is that through his research activities he develops enormous power for "sub-
sequent manipulation". In general, the most serious issues that seem to arise
in protection of the subjects of research in the social sciences are of two
sorts. The individual research subject may risk violations of confidentiality
and privacy. The community of which the subject is a part is also put at risk
as the social scientist uses the information in his publications or as the basis
of his consultative opinion leading to the formulation of public policy. A de-
tailed analysis of the risks and benefits of social science research will be
the subject of another staff paper.
Boundaries between research and practice in education and welfare :
It is of historical interest that most federal regulation concerned with
protection of human subjects of research has been concentrated in the Department
of Health, Education and Welfare. In fact, PL 93-348 requires of the Commission
that it make recommendations to the Secretary, DHEW. Curiously, thus far almost
all regulation oriented toward protection of human subjects has focused on health.
There seems to be little recognition in existing legislation or regulation that
considerable research involving human subjects is done in the fields of education
and welfare.
Experiments in education and welfare may influence the lives of "experimental
subjects" dramatically. The risks and benefits are as least as great--though not
often so immediate- -as those conducted in the field of health. Here, in contrast
1-24
to the social sciences, there is virtually always a problem in distinguishing
between research on one hand and, on the other, accepted and routine "practices".
Thus, if one wishes to consider the boundaries between research and practice in
these fields one might begin with essentially the same conceptual model developed
in this paper.
The Commission is aware of the problems in education and welfare but has--
for the time being- -decided to consider these problems outside the "boundaries"
of its charge.
1-25
ADDENDUM - September 24, 1975
Elaboration of definitions and concepts
One purpose of this addendum is to clarify the definitions of some
of the words used in the paper, "The boundaries between biomedical or
behavioral research and the accepted and routine practice of medicine
"July 14, 1975)--hereafter referred to as "this paper". A second purpose
is to clarify and formalize some of the concepts that were developed in
this paper. Special attention will be devoted to those words and con-
cepts that provoked either controversy, misunderstanding, or both in the
course of discussion of this paper at the September, 1975, meeting of the
Commission.
The definitions of words used in this paper are those contained in
Webster' s Third New International Dictionary . The words that presented
the greatest problems are: research, experiment, practice, intent, and
innovative. Examination of their definitions in the dictionary will re-
veal that several have many definitions and when used in various contexts
may have differing connotations. An attempt has been made throughout this
paper to adhere to dictionary definitions so as to avoid development of un-
necessary neologisms.
Dictionary definitions
Research : The most important definition of research used in this
paper is that provided for the word when used as a noun. "2a: Studious
inquiry or examination; especially: critical and exhaustive investigation
1-26
or experimentation having for its aim the discovery of new facts and their
correct interpretation, the revision of accepted conclusions, theories, or
laws in the light of newly discovered facts, or the practical applications
of such new or revised conclusions, theories, or laws."
Experiment : Noun, "la: A test or trial; b(l): A tentative procedure
or policy — especially : One adopted in uncertainty as to whether it will
answer the desired purpose or bring about the desired result... c: An act
or operation carried out under conditions determined by the experimenter
(as in a laboratory) in order to discover some unknown principle or effect
or to test, establish, or illustrate some suggested or known truth." As an
intransitive verb: "To engage in experimentation."
Experimental : Adjective. "1: Of, relating to, or based on experience.
2a: Founded on, derived from, or discovered by experiment (the heart of the
experimental method is the direct control of the thing studied)."
Practice . Noun, "lb: Actual performance or application of knowledge
as distinguished from mere possession of knowledge: Performance or application
habitually engaged in; usually: repeated or customary action. c(l): The
usual mode or method of doing something (the practice is to use a local anes-
thetic)." As an intransitive verb: "3: To exercise or pursue an employment
or profession (as medicine or law) actively. Synonyms: Act, operate, proceed."
Intent : Noun: "la(3): The state of mind or mental attitude with which
an act is done: volition--b: An end or object proposed: aim." As an ad-
jective: "2b: Having the mind or will concentrated on some end or purpose:
determined, resolved, bent."
Innovative : Adjective: "Characterized by, tending to, or introducing
innovations."
1-27
Innovation. Noun. "1: The act or an instance of innovating: the
introduction of something new; 2: Something that deviates from estab-
lished doctrine or practice: Something that differs from existing forms."
Discussion
Much of the discussion will focus on the necessity to use intent as
a taxonomic device to distinguish research from practice. This will have
the effect of focusing further discussion on this apparently controversial
term. However, on page 5 it is pointed out that intent is only one device.
It is the one that is traditionally used by the individual professional;
this will be discussed further below. The second system of classification
is a social system; ie, that system generally employed by groups. These
groups- -such as Peer Review Groups, Institutional Review Boards, and Re-
gulatory Agencies make their distinctions on the basis of whether an
activity or procedure meets certain criteria of acceptance or approval.
Some of these groups also have the authority to either not accept or dis-
approve. Thus, a group may first classify an activity as either research
or practice and then further proceed to say it is either bad research or
bad practice and therefore either unacceptable or unapproved.
These two systems of classification are almost never totally dissociated
in the conduct of activities in the real world. Thus, even when intent is
used as the taxonomic device (page 6) the definition of practice contains
the concept that the physician uses "...what he considers to be the best
available technique and technology...". This, then reflects the judgment
of the individual professional as to what is customary, standard, accepted,
1-28
approved, and so on. Subsequently, throughout the paper, examples are
given of situations in which the judgment of the individual may conflict
with that of a Peer Group or Regulatory Agency. It is in these situations
that some of the more serious boundary disputes are enacted.
Why is the device of intent necessary ?
Examination of the dictionary definition of research reveals that a
key word in the definition is "aim". In the dictionary this same word
is equated with one important definition of intent. The dictionary de-
finition of practice is not particularly helpful in these considerations.
It does seem to support what has been described previously as the social
or group approach to defining practice. Thus, when practice is defined as
a "repeated or customary action" it seems to relate to the standards of a
community of professionals. Some support for using the device of intent
to identify practice may be found in the Oath of Hippocrates: "I will
follow that system of regimen which, according to my ability and judgment,
I consider for the benefit of my patients..." The restatement of this
Oath at Geneva in 1948 includes the following statement: "The health and
life of my patient will be my first consideration." Thus, it seems appropriate
to equate such words as contained in these avowed purposes of the medical pro-
fession with the dictionary definition of intent.
Admittedly, there is no way to measure or evaluate intent unless it is
expressed clearly to the individual or group having the responsibility to
evaluate or measure it. Ordinarily, this will require that a formal plan
(protocol) be developed as discussed on page 7. However, some further
support for using intent as a classifying device may be found by reexamining
1-29
the purposes of this paper (pages 1-3). In order to respond to any of
these purposes it will be necessary to have intent described as part of
a formal plan. Clearly, most activities of health care professionals are
not conducted according to formal plans. At the moment a decision is made
to proceed with an activity--be it complex or simple--the "plan", if any
first exists only in the mind of the professional. Before the plan exists
the professional ordinarily first perceives his intent. His intent may be
to practice or it may be to conduct research or it may be to accomplish
both intents simultaneously. Existing customs and regulations demand that
if his intent is research it should ordinarily be expressed formally as a
protocol and reviewed by an IRB before he proceeds. Further, in the ne-
gotiations between the professional and the individual (patient or subject)
the intent of the professional should be made clear. The professional should
either agree (often tacitly) to the requeat of the individual to establish
a physician-patient relationship or he should invite the individual to play
the role of subject either in addition to or in place of the role of patient.
Part of informing for purposes of securing consent in the research context
is to advise the prospective subject that--to some extent at least--the pur-
pose (intent) of the activity will be to develop new knowledge.
The definition of research provided in this paper is designed, in part,
for the benefit of the professional who will wish to distinguish which of his
activities may be viewed (by others) as research. He may be advised that, at
some moment when he is considering performing some activity, he can consider
whether his intent is in part or in whole research a3 contrasted with practice.
1-30
In that case he may be advised further to express his intent in the form
of a protocol and have it reviewed by an IRB. He may also be advised to
conduct his consent negotiations with the prospective subject so as to
make clear his intent to that individual.
Definition of research
The definition of research abstracted from the dictionary is a per-
fectly adequate description of what "good" research ought to be in general.
However, for the purposes of these considerations the dictionary definition
is not adequate to define or consider the boundaries between research and
practice. It is not adequate in at least two respects. Firstly, much of
what all will agree are part of the customary "practice" activities of
physicians fulfills the requirements of the dictionary definition of research.
Thus, for example, most diagnostic activity might technically be classified
as research. A powerful argument has been developed by Francis D. Moore
(Daedalus, Spring, 1969, cited in detail below) that even much therapeutic
activity performed by a physician might be considered "experimental". In
this regard the dictionary definition is too broad to meet the purposes des-
cribed on pages 1-3.
On the other hand, the dictionary definition may be considered too narrow
to meet these same purposes. The dictionary definition as mentioned earlier
is a good general definition of "good" research. However, a strict interpre-
tation of this definition would exclude "bad" (poorly designed) research.
Certainly, one would wish to have such activities reviewed by an IRB.
Now let us reexamine the definition of research presented on pages 6-7.
The basic operational elements of the dictionary definition have been retained.
1-31
What has been removed are the words such as "critical", "exhaustive", and
so on, that have connotations of quality. The words "investigation" and
"experimentation" have been replaced with "manipulation observation, or
other study". As may be seen in the dictionary definitions of the words
experiment and experimental, these words—when used to establish a subset
of the universe of activities which may be classified as research- -have
connotations of intervention, manipulation, or control over the thing or
situation being studied. The word manipulation is meant to incorporate
most research activities that can be defined as experimentation. It would
be perfectly appropriate to replace the word manipulation with experi-
mentation if it is clearly specified which definitions of this word are
intended. The first two definitions of experiment as a noun would in-
corporate the concept of innovative therapy (to be discussed subsequently).
It is intended in this definition to incorporate innovative therapy as a
form of research.
The purpose of the wording "--or of anything related to that human
being that might subsequently result in manipulation of that human being--"
is to recognize that some research that is not performed directly on human
subjects may put these subjects at risk. Thus, as discussed elsewhere in
this paper, examination of medical records may put humans at risk; it may,
for example, lead to invasions of privacy. This concept is spelled out in
greater detail in the second sentence of the definition.
Also retained from the dictionary definition is the concept of "aim".
1-32
Here it is expressed as intent. The reasons for this have been described
in detail.
The final concept retained from the dictionary definition is that "of
developing new knowledge". However, for purposes stated earlier, this brief
phrase has been stripped of those descriptors calling for quality. Also, to
distinguish this from customary diagnostic and some therapeutic activities
the words have been inserted: "and which differs in any way from customary
medical (or other professional) practice".
Various suggestions were made by several Commissioners to add additional
attributes to the definition of research. Thus, for example, it was suggested
that research might be an activity so designed as to lend itself to statistical
analysis; be replicable, and so on. For the same reasons the decision was
made to remove from the dictionary definition terminology having qualitative
connotations, I should like to avoid adding these attributes. Additionally,
some of these attributes would exclude some sorts of activities discussed on
pages 7-8 in the two paragraphs concerned with problems in developing formal
protocols.
It was suggested further that research might be distinguished from
practice by virtue of examination of the fiduciary relationship or the
social structure of the relationship between the professional and the in-
dividual (patient or subject). Similarly, the reward system- -whether
payment is made to the individual for participating in the activity- -might
be considered a classifying device. I suggest that these are not classifying
devices; they are rather phenomena that occur as a consequence of having made
the classification.
1-33
Innovative therapy reexamined
The development of innovative therapy as a distinct category of the
activities of health care professionals was established firmly by Dr.
Francis D. Moore in his article entitled "Therapeutic Innovation: Ethical
Boundaries in the Initial Clinical Trials of New Drugs and Surgical Procedures"
(Daedalus , Spring 1969). The title of this essay implies that Dr. Moore was
concerned only with therapy. However, he clearly calls attention to the fact
that this classification includes prophylaxis (eg, oral contraceptives and
vaccination are used as two of his major examples) and "...the initial em-
ployment of untrained personnel in the care of patients." The class he
develops is somewhat analogous to that defined as "therapeutic research"
(Research on the Fetus; Report and Recommendations , page 6) including not
only treatment but also prophylactic and diagnostic methods. Thus, perhaps
a better name for this class of activities might be Innovative Practice.
In defining this class of activities it seems appropriate to use the
definitions of innovative and innovation found in the dictionary. In this
case it should be made clear that "the introduction of something new" does
not only mean introducing a modality that is new in kind. When an old
modality is used for a new purpose this may be classified as innovation.
For example, aspirin, which has been used for many years to provide sympto-
matic relief for pain, inflammation, and fever, has recently begun to be
used as an anticoagulant. Similarly, thyroid ablation by either surgical
removal or administration of radioactive iodine was introduced for treatment
of hyperthyroidism. Performing the same procedure for purposes of relieving
angina pectoris must be considered an innovation. Additionally, using a higher
1-34
or lower dose either of a drug or of radiation administered for therapeutic
or diagnostic purposes must also be considered innovative.
It is proposed that innovative therapy (or practice) be developed as
a distinct subset of the activities of physicians and other health care
practitioners. It may be distinguished from practice (as defined in this
paper) by virtue of its newness. The attributes of the activity which may
be new include any or all of the following: It may be an entirely new
modality or it may be an old modality used either for a new purpose, in a
new dose schedule, or in combination with other new or old modalities. In-
novative therapy (or practice) may be distinguished from other sorts of re-
search in that there is some scientific basis for predicting that some direct
health benefit will accrue to the subject. Thus, innovative therapy most
closely approximates the category of "therapeutic research" as previously
defined by the Commission. In general, innovative therapy should be con-
ducted and reviewed as if it were research. Some possible exceptions will
be explored below. Operationally, the definition of innovative therapy
should accommodate the knowledge that many aspects of the practice of
medicine can be stated in only general terms; eg, how in general one might
diagnose or treat a category of patients. As one examines most diagnostic
and many therapeutic activities conducted on a particular patient, a rigorous
analysis of the activities will indicate that something that someone might
call innovative has been performed. Thus, the physician commonly "experi-
ments" with various doses of various drugs to find the combination most
suited to the particular patient. The surgeon and the psychotherapist also
1-35
commonly deviate from textbook descriptions of their particular types of
therapy in order to be responsive to the needs of the particular patient.
Therefore, the definition of innovative therapy- -for purposes stated on
pages 1-3- -should assume that the adjective substantive will be applied to
the word innovation.
In the original draft of this paper it was proposed that areas of in-
novative therapy that should be scrutinized particularly carefully include
the following: surgery, invasive diagnostic techniques (including biopsies
and catheterizations), diagnostic and therapeutic devices, radioisotopes,
radiation and other physical forms of therapy, behavioral approaches to psy-
chotherapy, and the development of new health professionals. There should
be some sort of group mechanism to establish standards of acceptance and
approval for these categories of activities. Activities in many of these
categories are being defined by the individual professional as either re-
search or practice purely on the basis of his intent. What is needed is
some sort of group to apply the classification devices of acceptance, ap-
proval, safety, and efficacy. An appeal for the development of such standards
for behavioral approaches to psychotherapy has recently been transmitted to
the Commission in a letter from Dr. Daniel X. Freedman to Commissioner Brady
(July 3, 1975).
In the initial draft of this paper it was proposed that- -to the extent
possible- -the FDA model designed to review the use of drugs be applied to
these other categories. Parenthetically, it is re-emphasized that this is
not necessarily to be construed as a requirement for new federal regulatory
agencies.
The FDA model--as a social device for developing standards of accepted
1-36
or approved (or safe and effective) --was discussed In this paper only to
the extent that It assists In defining the boundaries between research and
practice. However, it should be made clear that this model has additional
features which would be responsive to the wishes of the Commission as ex-
pressed at the September, 1975, meeting. The FDA model provides criteria
not only for defining which of the activities under its purview are research
and which are practice, It also provides criteria for determining which of
these activities are acceptable and which are not acceptable. A strong con-
cern expressed by some Commissioners is that some activities performed as
"therapeutic innovations" are considered by some to be practice activities.
Therefore, they may not have applied to them the social device of acceptance
or approval. Thus, "bad" therapeutic innovations may be introduced and may
cause inappropriate harm. In general, if for purposes of review, innovative
therapy (practice) is classified as research and reviewed by IRBs--and when
necessary, groups constructed to perform the functions of the FDA- -the likeli-
hood of introduction of "bad" innovative practices should be minimized.
The FDA model demands that certain criteria be met before a potential
therapeutic innovation might even be considered an acceptable research
activity. Thus, to the extent possible, the potential innovation must be
studied in appropriate animals. After appropriate animal and other pre-
clinical testing has been accomplished it may be deemed appropriate to pro-
ceed with phase I clinical studies. This is the point at which the inno-
vation has been classified as acceptable research on humans. Phase I clinical
1-37
studies on a drug are not truly innovative therapy. This is because the in-
tent of phase I studies is not to bring direct benefit to anyone; rather it
is to determine the toxicity (or relative safety) of the drug. In most cases
the subjects selected for phase I studies are normal. Innovative therapy
ordinarily begins with phase II in which--in a very controlled setting—the
drugs are first administered to individuals having the disease which the
drug is designed to treat (target population).
Let us now examine this part of the FDA model to see how it might be
applied to other categories of innovative therapies (practices). First, it
should be recognized that all concerned have agreed that in some drug studies
phase I should be omitted. This is particularly true when the drug to be
studied is one which on the basis of animal testing is predicted to have
serious toxicity and which is designed to treat a very serious disorder.
Thus, for example, some highly toxic drugs are designed to treat patients
with cancer. Customarily, clinical trials of these drugs begin in phase II;
ie, the first humans to receive these drugs after appropriate animal testing
are individuals with cancer in whom other forms of therapy have been tried
and have failed. It would be clearly inappropriate to expose normal subjects
to such high toxicities without the at least hoped for benefit of relief from
some serious disease.
Preclinical testing leading to the equivalent of a INDA clearly can
be accomplished with therapeutic innovations in many forms of surgery, in-
vasive diagnostic techniques (including biopsies and catheterizations), dia-
gnostic and therapeutic devices, and so on. Of the categories specified for
1-38
which the social device of acceptance or approval should be developed, the
only one in which preclinical testing may not be useful is behavioral
approaches to psychotherapy. Just as with drugs, preclinical testing should
allow us to predict to varying extents the safety and efficacy of these
other types of activities. There will be clear cases in which appropriate
animal models do not exist. However, this should be no more limiting in
most other categories than it now is with drugs.
Phase I studies as described for drugs would be clearly inappropriate
for testing of most other innovative therapies. As Dr. Moore has pointed
out, it is quite impossible to begin with small doses of surgery and gradually
build up until a toxic dose is found. However, there may be some therapeutic
innovations that we might wish to first test in normal subjects. Thus, for
example, in the development of new health professionals they might first
"practice" on normal volunteers. Similarly, in diagnostic radiology, an
important function of an innovative technique might be to distinguish
various types of abnormality from normal. In order to do this it is ne-
cessary to perform the new diagnostic radiologic technique on normal humans.
This presents little problem when the new technique involves producing a
roentgenogram in a new way delivering a low dose of radiation. On the other
hand, problems are presented when innovative diagnostic radiology is combined
with invasive (eg, catheterization) techniques or when high doses of radiation
must be administered.
In some other forms of innovative practices normal humans may be the
only intended target population. This is particularly true in considerations
of innovative prophylactic techniques such as immunization to various diseases
1-39
transmitted by viruses and microorganisms and in pharmacologic or surgical
approaches to contraception.
In considering the FDA model as it might be applied to innovative
therapies in general the most important phases in the development are II
and III. Phase II is begun after appropriate preliminary studies have in-
dicated that the hoped for benefit might be achieved and that the risks of
harm are likely to be outweighed by the probability and magnitude of the
hoped for benefit. In cases where it is impossible to make such determina-
tions or to develop such expectations based upon prior preclinical and phase
I testing there should be at least some theoretical framework within which
the innovation is first performed. It should be possible to state formally
within the context of some theoretical framework- -or, perhaps, by analogy--
that the criteria for beginning phase II have been met. As phase II is
begun, the innovation is performed in a stringently controlled environment
using a small number of subjects derived from the target population. After
sufficient data are accumulated to show promise of safety and efficacy in
this small target population, the activities proceed to be conducted ac-
cording to phase III standards. In this phase a much larger population is
exposed under much less controlled conditions gradually approaching those
under which the innovation might be used in practice.
The basis for developing innovative therapies in this fashion is as
follows: Very early in phase II the power to predict either the risks or
benefits of the innovation are minimal. With continuing experience and
repetitions of the innovation the power to predict both harms and benefits
increases progressively. The process of increasing the power to predict is
enhanced if the innovation is performed in a setting which most closely ap-
1-40
proxiraates the research (dictionary definition) setting. Thus, as one pro-
ceeds to attempt to benefit patients- -in accord with the definition of
practice as defined in this paper--one should ordinarily also be conducting
research.
In general, individuals conducting innovative therapy should be con-
sidered both health practitioners and investigators. The innovative therapies
should, in general, be conducted according to the highest standards of the
relevant scientific discipline. In that way it might be possible to avoid
the introduction of therapeutic innovations which have not been tested
adequately and wh ich in forma lfy pass into the realm of "usual and customary".
It is preferable to conduct formal and controlled experiments during the
process of innovation rather than after the innovation has become firmly
established as customary practice.
Once the innovation has become established as customary practice it is
much more difficult to study it in a systematic way to determine its safety
and efficacy. For example, in 1966, the National Academy of Sciences, in-
itiated- -under contract from the FDA--a review of drugs introduced into the
market between 1938 and 1962. The purpose of this review was to determine
whether each of these drugs was effective for the indications designated by
their manufacturers. Their report ( Drug Efficacy Study , Final Report to the
Commissioner of Food and Drugs, from the Division of Medical Sciences, National
Research Council, NAS, Washington, D.C., 1969) categorized drugs as effective
effective but, probably effective, possibly effective and ineffective; it
further made recommendations as to how data might be developed that would
lead to reclassification from intermediate categories to either effective or
1-41
ineffective. Now, 6 years later, the FDA is still proceeding with the
implementation of these recommendations. Many drugs originally placed
in intermediate categories remain on the market without having had the
deficiencies in documentation of their efficacy supplied.
If all innovative practices were studied systematically during the
process of innovation it is likely that fewer "bad" innovations would become
firmly established as customary practice. Had this been the usual approach
to innovation in the past, many individuals might have been spared sub-
sequently abandoned dangerous surgical procedures ranging from routine
tonsillectomies, to a variety of surgical procedures for treatment of
coronary artery disease, and to prophylactic portal-caval shunts for the
treatment of esophageal varices (a complication of cirrhosis of the liver).
In exceptional circumstances, an innovative therapy might be used in
a pure practice context. Thus, for example, a patient with a life- threatening
illness requiring emergency treatment may be delivered to a hospital that is
not prepared to perform the innovative treatment in conjunction with those
measurements that would be designed to contribute to the general fund of in-
formation on its safety or efficacy. Further, it may be that no alternative
therapy is available for this individual. In such situations it seems
appropriate to proceed with the new form of therapy as if it were "standard
or accepted". It further seems appropriate to proceed without having con-
sulted an IRB or any group that might impose group judgment on the activities
of the professional. However, in such situations, the physician (or other
health care professional) should be aware that subsequently he is likely to
be held accountable for his actions by some appropriate group. In cases of
1-42
this sort, the group is likely to be the hospital board. Perhaps in the
future it will be the IRB.
Conclusions
The definition of research presented in the first draft of this paper
i
should be retained. However, it should be pointed out that- -for reasons
described- -it departs to some extent from the definition provided by the
dictionary. Perhaps the Commission will wish to state in its recommendations
that this is not a definition of research; rather, this is a definition of
how a subset of the universe of activities of health care professionals may
be distinguished from practice. The purposes of making such a distinction
are at least those described on pages 1-3 of this (the original) paper. It
would be appropriate- -should the Commission wish--to replace the word, mani-
pulation (the first time it appears in the definition) with the word, experi-
mentation.
For practical purposes, the definition of research as provided in this
paper, includes innovative therapy (or innovative practice). This means
that any innovative practice in which the deviation from customary practice
is substantive should be conducted so that it most closely approximates the
standards of good research (as defined by the relevant scientific discipline)
without obstructing the intent to bring direct health benefit to the patient-
subject. It further means that the proposed innovative activity should be re-
viewed by an IRB, that the consent negotiation indicate that the activity is
being performed with--at least in part- -research intent, and so on.
The primary purpose of this paper was to consider the boundaries between
biomedical or behavioral research and the accepted and routine practice of
1-43
medicine. A reexamination of the paper indicates that the boundary that
presents the greatest problem is, in general, the boundary between accepted
and routine practice on one hand, and, on the other, innovative therapy
(or practices). The boundaries between practice and other types of research-
eg, what the Commission has previously defined as "nontherapeutic research"
will ordinarily present no problem whatever. The definition of research
provided in the dictionary can be applied without modification to almost
all "nontherapeutic research".
1-44
THE ROLE OF ASSESSMENT OF RISK BENEFIT CRITERIA
IN THE DETERMINATION OF THE APPROPRIATENESS OF
RESEARCH INVOLVING HUMAN SUBJECTS
Robert J. Levine, M.D.
October 27, 1975
The Commission is charged with the responsibility to consider, among
other things: "The role of assessment of risk-benefit criteria in the
determination of the appropriateness of research involving human subjects."
In the charge to the Commission a key word appears to be "role". The
charge does not seem to call for a comprehensive analysis of risks or benefits
but rather for an evaluation of what the role of such analysis or assessment
is or ought to be in the determination of the appropriateness" of research. Let
us begin with the assumption that we do not now knoi<7 what the role of such as-
sessment is. This should be discovered through the empirical studies of In-
stitutional Review Boards (IRBs) that are now being conducted by the Commission.
This paper will be more concerned with what the role ought to be.
In order to provide a basis for discussion of the role of assessment of
risk-benefit criteria it will first be necessary to agree upon what phenonomena
might be considered risks and what might be considered benefits. Thus, this
paper will begin with a general discussion first of risks and then of benefits
both to individual research subjects and to society (and various sub-sets
thereof). There will be no attempt to mention all conceivable risks and
benefits. Rather, there will be a classification of risks and benefits to
demonstrate in general what is meant when the words are used to describe or
predict the consequences of research.
After risks and benefits have been classified and discussed we shall
turn to the primary focus of this paper; ie, a consideration of the role of
assessment of risk benefit criteria. .. Implicit in the charge to consider the
role of assessment of risk benefit criteria is the suggestion that there
*These words, as used in this paper, are defined in the section en-
titled "Definitions of role and appropriateness".
2-1
might be other criteria which are or should be used to determine the approp-
riateness of research involving human subjects; this suggestion will also
be explored.
Many literature citations on risks and benefits are contained in: Katz,
J. : Experimentation with Human Beings , Russell Sage Foundation, New York,
1972. As a matter of convenience, this reference will be cited as often as
possible; such citations will appear as (Katz, pp. 10-12). Those who wish
to see the primary sources will find them cited on those pages. However,
since some novel examples have developed since 1972, it will be necessary
to cite other sources; whenever possible a survey article in a readily avail-
able publication will be used.
Definitions of risk and benefit
The definitions of risk and benefit to be used in this paper are those
contained in Webster' s New International Dictionary, Second Edition. Since
not all of the definitions contained in that dictionary are germane to the
present discussion, the following is a clarification of which definitions are
to be used.
Risk , noun: 1. Hazard; danger; peril; exposure to loss, injury, dis-
advantage, or destruction; as, mountain climbing involves great risks; risk
of assassination.
2. (This definition is not to be used). In forest-protection usage,
any agency that may cause a fire.
3. Insurance, a) The chance of loss or the perils to the subject matter
of the insurance covered by the contract; also, the degree of probability of
such loss, b) Short for the amount of risk, that is, the amount which the
2-2
company may lose, c) (Not to be used.) Loosely, a person or thing con-
sidered with reference to the risk involved in placing insurance upon him
or it.
The word risk will also be used as a transitive verb. 1. To expose to
risk, hazard, or peril; venture; hazard; as, to risk one's person cr reputation.
2. To incur the risk or danger of; to venture upon (something involving
risk); as, to risk a battle.
In this paper, risk will not be used as an intransitive verb.
Benefit, noun: 1. (This is an obsolete definition which will not be
used). A good deed.
2. (This definition has questionable application to the present dis-
cussion; in general the word as used in this paper is not intended to have
this definition.) Act of kindness; favor conferred; gift; benefaction.
3. Whatever promotes welfare; advantage; profit.
4. Specif.: a) Pecuniary advantage or profit.
5. 6 and 7 are either obsolete or not germane to these considerations;
they will not be used.
8. Pecuniary help in time of sickness, old age, loss of employment, or
the like.
Synonyms: Help, service, use, avail. Antonym: Injury.
Benefit will also be used as a verb, transitive: To be beneficial to; to
intransitive.
do good to; to advance; improve; profit. To gain advantage; to receive benefit;
to profit; as, he will benefit by the change.
The terms risk -.nd benefit as defined in the dictionary are not clearly
parallel constructions. Risk, whether used as a noun or a verb, is a term that
2-3
deals with prediction of some future occurrence of harm. Risk may be ex-
pressed in terms of probability that a certain harm will occur. The harm
or injury itself may be evaluated quantitatively; eg, it may be described
as either a large or a small harm. The meaning of such constructions as
"small risk" thus becomes unclear; this might mean a small probability of
an unspecified amount of harm or an unspecified probability of a small amount
of harm.
By contrast, the term benefit has no intrinsic connotations of prediction
or probability. Benefit, whether used as a noun or a verb, denotes something
of value that can be supplied upon demand or as one wishes. It is of interest
that the antonym of benefit, injury, is the phenomenon, the probability of
which we are stating when we discuss risk. It should be made clear at the
outset that while discussing benef its—particularly direct health benefits of
research—one is ordinarily discussing probability of hoped-for benefits.
It was observed by the Commission (transcript, September 13, 1975), that
the uses of the words risk and benefit, as defined in this section, are different
from those used in common parlance; this is intended. The author shares the
view that the common use of such shorthand expressions as "risk-benefit criteria"
or "risk-benefit analysis" has unfortunate practical consequences. The scholar
concerned with exact meanings of words finds the expression "risk-benefit" of-
fensive because it is dysmorphic; it seems to equate, or make parallel things
which really are not ( supra ) . Even more perilous is the effect such constructions
may have on the thought processes of decision-makers. Thus, it is possible that
decision makers at all levels (eg, Congressmen, Commissioners, IRB members, in-
vestigators, prospective subjects) may--by virtue of frequent exposure to the
2-4
expression "risk-benefit"- -come to think that these are, in fact, parallel
constructions. As they are determining the appropriateness of any particular
research proposal they may think that risks can be expressed in the same terms
(eg, units of measurement) as benefits and be misled to ill-founded decisions.
To avoid this problem, it is suggested that all communications describing risks
and benefits to individuals or groups having the responsibility to make decisions
be described in language that conveys the true meaning of what is to be assessed
and, perhaps, balanced.
There will be found throughout this paper various attempts to maintain
parallel constructions. Thus, for example, risk is parallel to probability
of benefit and benefit is parallel to risk manifested as harm.
RISKS
In this section there will be a consideration of the various sorts of
risks that are assumed by human beings as a consequence of research. It
should be understood at the outset that what is being categorized is risk-
not research. The categories in which risks will be considered are physical,
psychological, social, legal, and economic. These are the risks of research
generally. Research on human subjects has previously been classified (boun-
daries paper) as biomedical, behavioral, and social. Some examples will be
provided in this paper of risks to human beings that may occur as a consequence
of research not done on human beings; eg, basic laboratory research. Some types
of research present risks that are peculiar to that category of research. These
will be mentioned when appropriate. Otherwise it will be assumed that the reader
understands that certain sorts of research are much more likely than others to
present certain categories of risk; eg, interventive biomedical research is
2-5
likely to entail physical risk while social research is not.
Risks to subjects
Physical risks
In some types of experimentation with human subjects the physical risks
will be obvious. Thus, when a well-known technique is used with research in-
tent it should, in general, be relatively easy to provide a totally adequate
description of physical risk. A fully adequate description would include a
statement of the likelihood of its occurrence, its severity, its duration
after the research maneuver is discontinued, its reversibility, measures
that can be employed for its early detection, methods that can be employed
to treat or correct it, and so on (see section on adequate description of
risks and benefits). Thus, when coronary sinus catheterization (a standard
diagnostic maneuver) is done with research intent, all of the information
called for in the preceding sentence can be provided.
In some situations the physical risk assumed by the subject is somewhat
less obvious. Thus, for example, when a relatively new drug is used experi-
mentally in the treatment of a disease for which there exist other proved
drugs, one of the risks involved is that for the duration of the study the
subject will be deprived of the benefit of the proved drug. It may well be
that the experimental drug will prove to be superior to the proved drug; how-
ever, this cannot be assumed during the course of the experiment. Further,
the adverse effects of the experimental drug may prove to be either more or
less severe
2-6
than those of the proved drug. An extension of this problem is seen when
therapy is withheld either for purposes of doing placebo-controlled studies
or for the purpose of making physiologic measurements in patients with various
diseases without the influence of drugs.
Another form of subtle physical risk is that seen in "double blind" studies,
A subject in a double-blind" study may become ill and require the emergency
services of a physician who is unfamiliar with the study. It may be necessary
for that physician to implement treatment promptly. It may be impossible to
implement rational treatment without knowledge of what drugs the subject (now
become patient) is taking. Thus mechanisms must be developed to assure im-
mediate access to the codes for "double-blind" studies in emergency situations.
In fact, such mechanisms have been developed and are functioning well. A de-
tailed account of these will be provided should the Commission wish it.
The physical risks of some forms of experimentation are totally unknown.
This is the case in most "phase-1" drug studies as well as in most other
studies when an intervention is performed on human beings for the first time.
some
Ordinarily in these cases there is background of experience derived from ex-
periments on animals which will help predict with varying degrees of con-
fidence what the risks might be to humans. However, it must be clearly under-
stood that one never knows what the adverse (or for that matter, beneficial)
effects of any intervention in humans will be until the intervention has been
adequately tested in humans. This statement may be extended further: The
effects of any intervention in any particular class of humans cannot be known
with certainty until that intervention has been adequately tested in that
particular class of humans. For example, as the Commission has previously
2-7
considered in great detail, some drugs that are acceptably safe for adult
males produce severe adverse effects when they are administered to pregnant
women or infants. In fact, it is through adverse reactions to various in-
terventions that we have discovered that some "classes" of humans exist.
Thus it was through thorough investigation of certain individuals with drug-
induced hemolytic anemia that we learned of the existence of glucose 6-phos-
phate dehydrogenase deficiency. Similarly, through adverse reactions to suc-
cinylcholine we learned of plasma cholinesterase deficiency. These 2 con-
ditions are genetically determined enzyme deficiency states that have no
clinical manifestations unless the afflicted individual is exposed to some
drug or other chemical which precipitates reactions peculiar to these in-
dividuals. The reactions may be severe; in fact, at times, lethal.
In some other cases physical risks might be totally unknown while an
intervention is in its experimental phases owing to long periods of latency
between the intervention and the development of the adverse effect. For
example, the suggestion that reserpine used in the treatment of hypertension
might increase substantially the risk of women to develop breast cancer was
first made approximately 20 years after this drug had become an "accepted"
(non-experimental) form of treatment for hypertension.
Some physical risks may be derivative. For example, innoculation of
children with rubella may result in inadvertent transmission of that virus
to a pregnant relative, neighbor or teacher. This, in turn, if the pregnancy
is in the first trimester, could result in the birth of a baby with serious
deformities.
Psychological Risks
It is important to distinguish between the psychological risks of re-
search in general and risks that are peculiar to psychological or other be-
2-8
havioral research.
Let us first consider some of the risks that are more or less peculiar
to behavioral research. First, it should be acknowledged that much research
directed at the patient (subject) with a behavioral disorder carries with it
risks that have been classified as physical, social, legal, and economic. Thus,
for example, early trials of electric shock therapy were associated with a high
incidence of serious injuries to the spine. Many of the drugs used to treat
psychiatric disorders present to the subject the risk of physical harm; con-
spicuous examples include severe liver degeneration--often fatal--as a con-
sequence of administration of iproniazid (an antidepressant no longer in use)
to depressed persons and obstructive jaundice in association with phenothiazine
(a class of major tranquillizers) administration. Other risks have their
parallels with those described for physical risks but which may be manifest
as psychological risks in the context of research directed at psychiatric
patients. These include the consequences of withholding of "proved" therapy
in order to test experimental therapy, placebo-controlled studies, and "double-
blind" studies. It is even more difficult in this category to predict either
the adverse or beneficial effects of previously untested interventions owing
to the fact that many human psychological disorders have no appropriate animal
analogs.
Some special problems are presented by psychological research in which
one knowingly invades the psychologic integrity of the subject. Thus, for
example, in the well publicized studies of Milgram (Katz, pp. 358-365), de-
signed to test some conditions of obedience and disobedience to authority,
normal volunteers were deliberately deceived. They were led to believe that
at the command of the investigator they were inflicting pain upon third persons.
2-9
In some cases they were led to believe that the pain they were inflicting by
way of electric shock was unbearable. Some of the subjects experienced ex-
treme anxiety. Some were startled to learn the brutality of which they were
capable; some detached themselves from a sense of responsibility by persuading
themselves that they were not, in fact, making the decision to inflict pain;
they were merely acting as agents of a superior decision-maker.
In studies by Bressler, B. et al (Katz, pp. 365-369) traumatic neurosis
was deliberately induced by placing normal subjects in an environment of
sensory deprivation. Some of the subjects manifested lingering disturbance
after the experiments terminated. The investigators concluded, among other
things: "Although the problems indicated were naturally latent, we believe
they would not have emerged with the same intensity in ordinary circumstances."
Some experiments with hallucinogens and psychotomimetics have produced
prolonged and, at times, serious behavioral aberrations. The risks and
benefits of psychosurgery will be the subject of another paper.
There are also a variety of derivative psychological risks. For example,
the successful treatment of an alcoholic or a neurotic may--in the view of
his or her spouse- -change that individual to someone whom the spouse "no
longer knows". This, in turn, may precipitate divorce or various other
personal dislocations.
Many investigators pre-screen prospective subjects to see if they are
"fit" for participation in the research. Some of the effects of this pre-
screening for psychological studies may be quite traumatic. For example, in
one study prospective subjects were informed that the investigators wanted
to show a stress-provoking movie to "normal" individuals to determine their
2-10
physiological reactions to the stress. They were further informed that
prior to viewing the movie they would have some standard psychological
screening tests. After these tests some were informed that the investigators
had decided- -based on the results of the tests--not to proceed to show them
the movie. One can imagine the reaction of these prospective subjects who--
by implication- -were informed that they had been found psychologically unfit
to see a movie.
The phenomenon of rejection based upon "failure to pass" pre- screening
tests for "normality" is not limited to psychological research. Many persons
who assume they are healthy learn through such tests that they have various
chronic diseases. Is this a harm or a benefit?
Many types of biomedical research interventions that are not directed
at the subject with a known behavioral disorder have proved, over the years,
to present serious psychological risks some of which were reversible and some
of which were not. Thus, it was not anticipated that the administration of
reserpine for purposes of treating hypertension would result in a significant
number of individuals in a rather precipitous development of severe- -and often
agitated—depression; some of these individuals proceeded to commit suicide
before the physician-investigator was even made aware that depression existed.
Similarly, during the early developmental stages of palliative surgery for
Parkinson's Disease and cerebral angiography some individuals were left with
serious, and often irreversible, behavioral deficits.
Serious psychological problems are occasionally presented to certain
susceptible individuals just by virtue of being made aware that they are
research subjects. Some people find it quite stressful to know that they
2-11
are venturing into the unknown; this is particularly true in situations
in which in order to become research subjects they must agree not to avail
themselves of "established" diagnostic or therapeutic alternatives. The
stress they experience may bemanifestedas any of a variety of behavioral
aberrations or psychosomatic complaints.
One psychological reaction that arises fairly commonly in relation to
(self-blame) .
agreement to participate as a research subject is guilt. When an individual
consents to research upon his own person there may be some motivating factors
in his agreement to become a subject which may or may not be discussed frankly
during the consent negotiations. Thus, at a time when "standard and accepted"
therapy for a melanoma on the thigh was considered amputation of the entire
hindquarter, an avid golfer who was invited to participate in a comparative
trial of local excision might have a problem. He might be motivated to par-
take of this relatively untested therapeutic approach largely because he
wanted to continue to play golf. On the other hand, he might wonder whether
he was behaving responsibly in relation to his family and friends by agreeing
to take what might prove to be a greater risk of dying of the disease. A
somewhat analogous situation might be seen in a woman with breast cancer who
agrees to participate in a clinical trial of "lumpectomy" at a time when- -as
she understands it--the standard and accepted practice is radical mastectomy.
The primary and perhaps undisclosed motivation for her consent is cosmetic.
Even if there is relatively no guilt at the outset there may be severe
guilt reactions if the risk actually becomes manifest as harm. Thus, if the
individuals mentioned in the preceeding paragraph subsequently are informed
2-12
that they have metastatic tumors they may develop the belief—which might
or might not be appropriate- -that if only they had chosen to proceed with
the standard and accepted therapy they would not be in their present pre-
dicaments. It is an easy step from there for the subject to equate his de-
cision with one to commit suicide. In these cases the guilt reaction becomes
manifest only if the cancer recurs. The very same individuals might be quite
pleased with themselves if five years later there was no evidence of recurrent
tumor and they were much less incapacitated or deformed by the "experimental"
surgery than they would have been by "standard and accepted" therapy.
This potential for guilt occurring as a consequence of an experimental
failure is, of course, not easily predictable. To cite two studies
the outcomes of which are now known: The parents who agreed to have cortisone
(standard and accepted therapy for acute rheumatic fever at the time) withheld
from the treatment of their children with acute rheumatic fever subsequently
must have been very pleased to learn that the study proved that the hormone was
ineffective for that indication; their children received the benefit of not sus-
taining all of the adverse consequences of cortisone administration. They must
feel that they made a "good" decision. On the other hand, consider the study
in which antibiotic treatment for streptococcal pharyngitis was deliberately
withheld. The young men who developed glomerulonephritis (a serious and some-
times lethal kidney disease) probably felt that they made a "bad" decision. They
are probably not particularly relieved to learn that they contributed to our
understanding of which strains of the organism are most likely to produce glomeru-
lonephritis.
2-13
The problems of guilt on the part of the consentor when risk becomes
manifest as harm are probably even more serious in situations in which
proxy consent was given.
In some situations the process of negotiating informed consent may
be quite threatening. A physician-investigator may approach a
patient with cancer for purposes of asking him to participate in a clinical
trial. The investigator may assume incorrectly that the patient "knows" he
has cancer because his personal physician has informed him. Thus during
the process of informing for purposes of soliciting consent the patient may
be deprived of the defense mechanism of denial.
In some situations—in the judgment of an IRB--the consent negotiation
has been even more risky than the actual proposed research. An example of
this was a proposal to approach individuals with suspected or proved acute
myocardial infarctions shortly after their admissions to coronary care units
with a proposal to do research that bore modest risk. It was the judgment
of the IRB that the anxiety that might be produced in the patient by sub-
jecting him to the necessity to make such a decision might be substantial.
Conceivably- -although this cannot be proved- -the degree of anxiety that might
be produced might contribute to the development of further damage to the heart
or to the development of "coronary care unit psychosis".
Distrust is another reaction that can be precipitated in individuals by
research. The decision to classify distrust as a psychological risk to the
subject is arbitrary. This phenomenon has ramifications in the categories of
2-14
social risk to individuals as well as risks to society.
Let us first consider how distrust might develop in the individual as
he is being asked to consent to become a research subject. The individual
might perceive, often quite correctly, (cf boundaries paper) that the role
he is being asked to play (subject) differs in some important way from the
role he expected to play (patient). Thus, to some extent, the professional
with whom he interacts might see the development of new knowledge as being
an important goal which might compete with the goal of proceeding most
efficiently to restore the health or well-being of the patient. Some sub-
jects verbalize this as being used as guinea pigs.
In most consent discussions it should be made clear to the prospective
subject that he has the right to refuse to become a subject. In fact, when
the subject has a relationship with the investigator which has any potential
for coercion there is usually added to the consent form a "non-coercive dis-
claimer". Such relationships might include physician-patient; employer-em-
ployee; faculty-student; and so on. When such relationships exist it is
ordinarily stated on the consent form (or in the consent discussion) that
the prospective subject is free to refuse to participate in the research and,
further, that he is free to withdraw from the research at any time. Further,
there is an assurance that such refusal or withdrawal will in no way adversely
prejudice his future relations with the investigator or the institution. Thus,
for example, if he is a patient, refusal to become a subject will not in any
way adversely prejudice his status as a patient. One wonders how often a
patient upon reading such language questions (perhaps not aloud) whether
the physician- investigator is capable of experiencing rejection (refusal
2-15
to consent to become a subject) without retaliating in kind; ie, rejecting
the patient in some way by becoming less responsive to his demands or wishes.
Thus, a subject who has not been coerced might feel that he has or might
have been. This, in turn, sows the seeds of a breakdown in trust in the
relationship between the professional and the individual.
Similarly, the knowledge that research is being done in an institution--
eg, a university hospital--may create in the community a sense of distrust.
In many communities it is "general knowledge" that if you go to a university
hospital you will be used as a "guinea pig". It is not at all uncommon that
people who are asked to become research subjects in a university hospital ex-
press surprise. They say, for example, that they didn't know that they would
be asked and offered the option to refuse. They just assumed that research
would be done on them- -perhaps without their knowledge. The risk
of this sort of misunderstanding is that some individuals may elect not to
go to a hospital when they feel sick and this decision may be detrimental to
their health.
Social risks
Social risks are inextricably entangled with issues of confidentiality.
Thus, through research an individual may become identified as a drug or
alcohol abuser; as a participant in various deviant sexual practices; as
having any of a variety of diseases which may be deemed unacceptable by
his family, social or political group; as having a higher or lower income
than his colleagues might have predicted; as a felon; and so on. If such
information becomes known to the wrong individuals it might cost the subject
his reputation, job, social standing, credit, citizenship, and so on. Ob-
2-16
viously, this issue extends into the categories of legal, economic, psy-
chological, and (rarely, one hopes) physical risk.
A somewhat less obvious social risk also involving issues in con-
fidentiality and privacy is presented by the mode of access used by some
investigators to prospective subjects. The investigators may attempt to
secure the names of prospective subjects through hospital records, school
records, social service agencies, and so on. Thus, when the prospective
subject is contacted directly by the investigator, he may wonder what other
aspects of the presumed confidential relationship had been violated. Such
practices tend to undermine the confidence of individuals in health delivery,
social service, and educational systems.
Legal risks
Legal risks are presented to subjects of social, behavioral and biomedical
research. A majority of the legal risks might be viewed as a subset of social
risks. Thus, through violations of confidentiality the subject may become
liable to a variety of civil or criminal actions. The legal risks to the
subject of social and behavioral research will usually be obvious to all con-
cerned. The legal risks of some biomedical research may be more subtle (see
social risks).
In some states (eg, Ct.) there are statutes which grant testimonial
privilege as well as confidentiality to the psychiatrist-patient relation-
ship. Others (eg, 111., N.C., La., Pa., N.Y.) grant sharply limited testi-
monial privilege to the physician-patient relationship. Although no state
grants privilege to investigator-patient relationships, there are federal
research shield statutes providing limited protection particularly for
patients and subjects in drug and alcohol abuse programs. Thus, there is
the potential problem of law enforcement agencies gaining access to lists
of names of known felons through subpoena of the records of an investigator:
2-17
eg, a clinical chemistry laboratory, a methadone maintenance program, or
a criminologist.
Since a comprehensive review of the law as it relates to research is
about to be published there will be no further discussion here. This book,
entitled The Risks of Social Research , edited by Paul Nejelski and Batya
Miller of The Institute of Judicial Administration, New York City, includes
not only the papers presented at The International Conference on Research in
Conflict with Law and Ethics (March 1974) but also a comprehensive review of
statutes, regulations, and court cases in the field. The Commission may wish
to examine this book prior to its publication.
Economic risks
Subjects may be called upon to pay for procedures that are undertaken
with research intent. This is a particular problem in complex activities
(cf , boundaries paper) with mixed intent in which it is not clear either to
the investigator or the subject which parts of the activities are practice
and which are research. Subjects may be called upon to pay for hospital
services which would not have been necessary had no research been done. Also,
subjects may suffer loss of income when they take time off from jobs to parti-
cipate in research; they may also have to pay baby-sitter fees, and so on.
A more serious economic risk is involved in the payment for damages to
the subject that may occur as a consequence of research. Since, as will be
discussed later, it is society in general that derives benefit from most re-
search, it seems appropriate for the economic risks to be distributed equit-
ably. This might be accomplished through development of adequate insurance
for research subjects. When research is sponsored by the government it may
2-18
be assumed that the benefits are in the public interest; in these cases one
might expect the government to subsidize the insurance premiums. Similarly,
when the research is sponsored by industry (development of drugs or devices)
one might consider asking the involved industry to subsidize insurance premiums.
This issue is discussed further under the heading of economic risks to society.
Through participation in research a subject may lose his insurability.
This is at least a two- fold problem. First, in the course of research various
diagnostic tests may be performed that, in the ordinary practice of medicine,
would not be considered indicated. These tests may reveal diseases or sus-
ceptibility to diseases which- -in the view of insurance companies- -would call
for either higher insurance premiums or refusal to underwrite life, health, or
disability insurance. A second part of this problem is that in many institutions,
research findings which are germane to considerations of health, are entered
directly into the subjects' (patients) medical records. When the subject sub-
sequently applies for insurance and signs a form allowing the insurance company
access to his medical record he generally has no idea as to the quantity or
quality of information thus disclosed.
Other risks with economic implications are discussed in the categories of
social and legal risks.
There are also a variety of derivative economic risks some of which are
obvious and some of which may be subtle. For example, some research is done
without the awareness of some third party who might subsequently be aske d to
share the economic burdens imposed by failure of the research to accomplish the
hoped for benefit. Thus, for example, in studies of post-coital contraceptives
an unaware third party (coital partner) might (post hoc) be called upon to share
the burdens (economic and decision-making) in regard to abortion or carrying a
potentially damaged fetus to term.
2-19
Risk s to society
The physical risks to society are, in general, those specified as de-
rivative physical risks to the individual. One might also consider the
physical risks to society that occur as a consequence of research which
does not directly involve human subjects. Thus, pure laboratory research
may develop an alien and particularly virulent strain of virus. An ap-
preciation of the magnitude of this potential problem may be gained by con-
sidering the introduction of measles into previously unexposed populations
by 16th century explorers. Dramatic fictional descriptions are contained
in Crichton's " Andromeda Strain " and Szilard's "My Trial as a War Criminal".
Here we might also consider the public health hazards inherent in disposal
of "nerve gas" and radiation wastes.
The potentialities of developing alien viruses and antibiotic resistant
bacteria are no longer the exclusive domain of science fiction. It was this
sort of potential around which there was much debate in 1969-1970 when James
Shapiro et al described their isolation of the pure lac operon ; in connection
with this controversy Shapiro quit science to enter politics ( Science 167: 964,
1970). Subsequently the controversy was renewed in connection with a recognition
on the part of molecular biologists of the potential biohazards of recombinant
DNA ( Science 185: 303, 1974). In the latter case, recognizing the grave
potential for development of oncogenic and other viruses, antibiotic resistant
bacteria, toxin producing potential of ubiquitous and ordinarily innocuous
bacteria, and so on, the molecular biologists called for a moratorium on many
aspects of research in this field. Another highly significant potential in
pursuing this line of research is that there is the possibility of either deliber-
ately or inadvertently altering the human genome. This latter possibility
2-20
has enormous ramifications that might more appropriately be discussed under
the category of risks to society.
Psychological risks to society are in general those discussed as deri-
vative psychological risks to the individual research subject; others were
presented as social risks to the individual- -the consequences of the two
types of violation of confidentiality.
Premature (or otherwise inappropriate) dissemination of either the
findings or the opinions of researchers may present psychological risks.
Thus, for example, phobias may be developed on a rather grand scale when
the public is informed that low-cholesterol diets may cause cancer; that
small breasts are associated with low I.Q.; that citizens of certain cities
(e.g. Dallas)are susceptible to violence by virtue of low lithium content
in the local water supply, and so on. On the other hand, false hopes may
be raised by premature or otherwise inappropriate dissemination of either
the findings or opinions of researchers. Consider, for example, how many
different surgical "cures" for coronary artery disease have been discovered
and abandoned in the past 20 years. Thus far we have discussed only the
activities of "legitimate" researchers; one might also consider the con-
sequences of public announcement of such "miracle-cures" as Krebiozin (for
cancer), copper bracelets (for arthritis), rainbow pills (for obesity), and
so on.
Social risks
It is often asserted that the major burdens of direct risks to subjects
are borne disproportionately by certain socio-economic groups. The Commission's
studies of IRBs should indicate whether these assertions are true.
Studies designed to compare certain social, ethnic, racial, religious, or
political groups may develop findings which--in the view of some members of a
2-21
group- -might have pejorative implications. Others in the same group may
view the same results as beneficial to the group. For example, when re-
search revealed that the incidence of suicide was much higher in female
physicians than in male physicians some female physicians found this to
be pejorative and supporting the "male-chauvinist" cause. Others welcomed
the information as beneficial supporting their efforts to develop constructive
affirmative action plans. Similarly, in a non-research context, some in-
dividuals perceive plans to offer abortion and sterilization services in
publicly- funded health care delivery systems as attempted genocide. Others
in the same groups perceive regulations to proscribe such activities as de-
priving them (as a class) of their rights to benefit from technology available
to others to aid them in what they consider rational family-planning.
Another line of research that presents considerable risk to society as
well as to certain individuals within society is that in which an attempt
is made to predict certain characteristics of individuals or group of in-
dividuals by genetic screening. Thus, for example, there is the recent re-
newed interest and controversy over attempts to link XYY chromosome patterns
with criminal behavior (Katz, pp 342-346). The unfortunate consequences of
this research and the publicity it received has been reviewed recently (Science
188: 1284-1285, 1975). Another related piece of research which has received
considerable publicity is the report that the same chromosomal markers can
be used to predict race and IQ; what was demonstrated was that there seemed
to be a correlation between one particular race and a low IQ (see for example
Lubs et al : Amer. J. Human Genetics 25: 47 A, 1974).
2-22
Another type of social risk is that some research either by its very
nature or by the sometimes callous manner in which it is performed may
shock the sensibilities of society (or some subsets thereof). Historically,
public sensibilities as they were perceived about 20 years ago, caused in-
vestigators who were then developing oral contraceptives to conduct the
early clinical trials in Puerto Rico. One doubts that this would be ne-
cessary today. In its time the "Kinsey Report" offended the sensibilities
of many citizens. At the time the "Masters and Johnson" work was being
publicized almost noone would have considered the "Kinsey Report" offensive.
Similarly, the second "Masters and Johnson" repoic shocked public sensi-
bilities at a time when the first report would probably have shocked very
few.
In some cases relatively little issue might be taken with the actual
research but rather with the apparently callous manner in which it was per-
formed. Thus, graphic descriptions of decapitation of dead abortuses have
shocked public sensibilities; possibly the very same research- -done with
greater sensitivity- -would have attracted little or no attention.
There are some sorts of risks to society that may be so large as to
constitute absolute barriers to proceeding with research. Whether or not
it is appropriate to incorporate these as risks for purposes of entering
them into the calculation of risk-benefit criteria will be discussed later.
One of the sorts of risks is presented by research that would change our per-
ception of what constitutes a person.
The heated controversy that can develop as to what constitutes a person
2-23
was recently well illustrated for the Commission in the debates about
fetal research. On one extreme there were ethicists and others who
firmly asserted that a person begins to exist at the time of conception
or, alternatively, at the time of implantation. On the other extreme
there were those that supported the concept that all of the attributes
that constitute personhood have not developed until approximately age
\\ to 2 years. A third point of view was that one might finesse the de-
bate by classifying personhood as a "premature ultimate".
Depending upon ones definition of personhood together with the prin-
ciple that a human person bears a value that takes precedence over all
other values it was asserted that under no circumstances could one deprive
a person of his life in order to achieve some goal (eg, a research objective).
Similar debates were conducted in the recent past about developing definitions
of death which would permit harvesting of viable organs or tissues for trans-
plantation purposes. It seems unnecessary to recount these debates here.
Similar debates have begun and may be expected to intensify as the
potentialities for genetic-modeling, cloning, hybridization with other
species, and so on become closer and closer to being technically feasible.
is presented by
Another sort of closely related risk to society actions which
A
may tend to erode the bonds of trust that one ordinarily assumes. This
has been discussed in other sections in relation to eroding the bonds of
trust between individuals and various institutions. In the research context
one is particularly concerned with the authority to give proxy consent. Thus,
as there is a growing public awareness about the possibility to give proxy
2-24
consent to risky research, there may be a gradual erosion of the bonds of
trust between those with limited capacities to consent on one hand and, on
the other, those authorized to--or even held responsible for--giving such
proxy consent.
The issue of distrust was discussed to some extent under the rubric
of psychological risks to subjects. It was pointed out that this issue
also had ramifications in the category of social risks to society. Thus,
for example, it was pointed out that in some communities it is "general
knowledge" that if you go to a university hospital you will be used as a
"guinea pig". In that situation—particularly when individuals are ad-
mitted to university hospitals- -the phenomenon of distrust can be- -to some
extent--dispelled through the process of informed consent.
Now, let us consider the social risks in cases in which truly informed
consent cannot be (or is not) secured prior to initiating the research. This
sort of research is common in some social and behavioral sciences. For ex-
ample, in some sorts of research—the so called "unseen observer" type of
research- -the experimental subjects are not even aware that research is
going on. A good illustrative case is presented by Humphreys "Tearoom
Trade" studies (Katz, pp 325 et seq) . This sort of research presents
several risks to society. In this particular case the "subjects" were
practicing homosexual acts in public restrooms. Such acts were at the
time classified as felonious in the city in which the research was done.
The risks to the individuals of violation of confidentiality have been dis-
cussed elsewhere. When the results of the research became publicized it is
quite likely that many homosexuals became distrustful of a society in which
2-25
they might be observed without their awareness. They must have been parti-
cularly frightened that they might be- -as a consequence of such observation- -
convicted of felonies. This type of "unseen observer" research has been re-
peated in many environments and there is a growing awareness of the possi-
bility that one might--at any time--be observed while engaging in activities
one might have considered private.
Other such studies have involved recording of the deliberations of juries
(Katz, p. 67 et seq) , and so on.
The "Tearoom Trade" studies also afford an opportunity to discuss two
additional types of risks to society which may occur as a consequence of con-
ducting social research. During the course of these studies the investigator
recorded the license plate numbers on the automobiles of the individuals whom
he was observing. Through contact with the motor vehicle registration auth-
orities he identified these people by name and address. He was then able to
make personal contact with them and then--with their informed consent --conduct
his in depth studies. However, it is conceivable that some individuals so
identified might have objected and might have developed the same sorts of
distrust mentioned earlier in relation to access problems when hospital, school,
or welfare records are used.
A second phenomenon observed by Humphreys and several other investigators
may be considered. When an investigator (or a type of investigation) becomes
well known it may be assumed on the part of some that the institutions or
groups of people he spends time with have a high incidence of the sorts of
people known to be of interest to the investigator. Thus, if Humphreys is
observed spending time at a certain institution it may be assumed by those who
2-26
observe him that that institution has within it a large number of homosexuals.
This becomes a particular problem when law-enforcement agents become aware
of the activities, interests, and whereabouts of certain investigators who
wish to study criminal behavior. Thus, for example, an investigator who
was known to be studying drug abuse patterns in a public high school would
tend to focus police attention on that high school and increase their sur-
veillance and questioning of students in that particular high school. Par-
enthetically, it might be noted here that some investigators have been threa-
tened with criminal liability if they did not reveal the names of those they
observed engaging in criminal activity.
In cases in which video-tapes are made of inaividuals engaged in what
they believe to be private acts the same issues of distrust are raised. How-
ever, in these cases protection of privacy becomes more problematic. It is
possible to remove an individual's name from a questionnaire. However, faces
and voices are sufficiently distinctive so that it may become impossible to
obscure the identity of those who were taped (video or audio) . In such cases
informed consent may be secured following the taping; the classic paradigm of
which most members of the public are aware is "Candid Camera". A prevailing
awareness that one might at any time be observed, recorded, or bugged, may
have a detrimental effect on the behavior of many individuals. An extreme
example of the potential impact of such activities on a society is described
by Orwell in "1984".
Some anthropologists by joining a society may actually change the nature
of the society. There may be considerable debate as to whether the changes
are detrimental or beneficial. An awareness of this potential is expressed in
2-27
the actions of the Phillipine Government to limit access to the Tasaday
Group.
Some additional discussion of the general risks and benefits to
society of social research was presented earlier in the "boundaries paper".
Legal risks
In the criminal justice system, society customarily views convicted
criminals as its adversaries. Thus, to the extent we develop research
shield privileges we may be imposing a risk on society. Also, one might
consider here the civil legal risks to the institutions that are performing
the research. For example, if an investigator i~ sued for malpractice (or
malresearch) while operating according to a protocol approved by an IRB, to
what extent do members of the IRB share the burden of this legal risk?
Economic risks
The actual cost of doing the research might be considered society's
"risk-capital" to develop information based on the assumption that the in-
formation will benefit society. Thus one criterion for determining the
appropriateness of the proposed activity might be: Is the information we
might develop worth that much money? For an interesting and provocative
analysis of how society responds to such questions see Calabresi, G. (Katz
pp 177-184).
Another type of economic risk borne by society is the cost of "taking
care of" individuals who are damaged by research. Here one is particularly
concerned with physical and psychological damage to research subjects. At
the very least, someone has to pay for the medical or psychiatric care ne-
2-28
cessary to treat the conditions induced by experimental interventions. An
even more serious (though presumably rare) threat is that of dealing with
an individual who as a consequence of research becomes permanently disabled.
Thus, through research, it is conceivable that one might take previously
normal and productive individuals and--by virtue of having induced paraplegia,
psychosis, drug-addiction, and so on--add them to the ranks of those who must
subsist on welfare.
The problem of insurance for research subjects is currently being studied
by the Secretary's (DREW) Task Force on the Compensation of Injured Research
Subjects. This task force is headed by Seymour Perry, M.D., Special Assistant
to the Director, NIH, in the Office of Program Planning and Evaluation, Office
of the Director, NIH. The orientation of this task force is to develop a
system of insurance that would compensate research subjects for harm that be-
fell them as a consequence of research. They are oriented toward de-
veloping the insurance on a "no-fault" or Workmen's Compensation Model. At
the time of this writing they do not seem near the point at which they might
make final recommendations. Apparently, they have encountered some difficulty
securing reliable data on morbidity and mortality rates for research subjects.
Perhaps the Commission would wish to establish formal communications with this
group.
Another approach to coverage has been developed at several institutions.
Thus, for example, the University of Washington (Seattle) has research insurance
underwritten by a private company (Aetna) . As I understand it this insurance
is based on a no- fault model and is developed along the lines of the state Em-
2-29
ployee's Compensation Act. One difficulty with this approach is that it
covers only individuals who are classified by state law as employable; ie,
children would not be covered.
Some other universities have their general malpractice coverage extended
to cover the activities of physicians who are engaged in research. In general,
this is a fault model. A claim must be made against the
physician-investigator and/or the institution; the settlement of the claim
is, in general, determined as in most malpractice claims.
Risks of not doing research might also be considered at this point. The
risks essentially will be a deprivation of society of the benefits to be
described subsequently. Here one is faced with the difficulty of anticipating
the outcome of future research. Thus the approach probably should be to
analyze what the risks might have been if some sorts of past research had not
been done. Examples that might be used include the development of various sorts
of immunization procedures (vaccinia, rabies, and polio provide dramatic ex-
amples), general anesthesia, antiobiotics, laparotomy, heart surgery, amniocen-
tesis, and so on. The "therapeutic-orphan" phenomenon has been discussed in
relation to research on the fetus and undoubtedly will be raised again in
considerations of research on children; it is a harm of not doing research.
BENEFITS
At the outset it should be made clear that there is no way to separate
the issue of quality of scientific design of research from the ethical con-
siderations as to whether it should be done. If research is badly designed
it is not likely to benefit anyone. Thus, it seems inappropriate to put human
beings at risk to develop information (or misinformation) that cannot con-
ceivably benefit either the individual or society. This principle is clearly
2-30
enunciated in the Nuremberg Code. The principle is extended in that Code
as follows: "The experiment should be conducted only by scientifically
qualified persons. The highest degree of skill and care should be required
through all stages of the experiment of those who conduct or engage in the
experiment." Principle 7 of the same Code requires that: "Proper preparations
should be made and adequate facilities provided to protect the experimental
subject against even remote possibilities of injury, disability, or death."
Subsequent commentators have extended this latter principle to mean that the
facilities available to conduct the research should also be adequate to assure
successful completion of the experiments. Thus, an important criterion of
whether it is appropriate to proceed with a particular research proposal is
that the experiment has been well-designed; the investigators who plan to
execute it have appropriate skill; and the facilities in which the experiments
are to be conducted are optimal not only to achieve the scientific objectives
but also to protect the rights and welfare of the subject.
Principle 3 of the Nuremberg Code reflects the context in which it was
written and clearly is not applicable to all research to be considered by
the Commission. "The experiment should be... based on the results of animal
experimentation...". This principle, in cases in which it is applicable, is
in part an issue in scientific design. The purpose of preclinical animal
testing is to develop information that permits better design of experiments
to be done on humans. Among other things, preclinical testing in suitable
animal models enhances ones power to predict the outcome of performing the
same intervention in a human. In some cases, the results of studies intended
2-31
to be preclinical will indicate that the experimental modality should not
be tested in humans.
One issue that may be separated from that of scientific design might
be stated as a corollary to parts of Principles 2 and 3 of the Nuremberg
Code: It is inappropriate to put humans at risk to gain information that
can be secured without putting humans at risk. This issue will be discussed
further in a subsequent paper.
Benefits to society
The discussion of benefits will be even more superficial and general
than that of risks. This reflects the bias of the author that those who re-
view research proposals to determine their "appropriateness" will almost always
find the benefits or hopedfor benefits clearly described. There is further a
tendency to overestimate the probability of achieving the hoped for benefit.
By implication it is suggested that risks tend to be underestimated. This
does not reflect dishonesty; rather it reflects the facts- -in the opinion of
the author- -that most investigators are focusing on the benefits and many
are either unaware of or inattentive to some of the phenomena described as
subtle risks.
The benefits to society will be discussed for 4 broad categories of
research: 1) Applied biomedical and behavioral research (performed with the
intent of developing or perfecting diagnostic, prophylactic and therapeutic
modalities) (this category is similar to innovative therapy (boundaries paper));
2) Basic research involving human subjects (done with the intent of improving
our understanding of the biology and psychology of normal and "abnormal" human
beings); 3) Basic research (not directly involving human subjects); and 4)
2-32
social research.
1) In general, the benefits of applied research involving human subjects
will be obvious. It is in the interest of society to develop improved tech-
niques and technologies for the diagnosis and treatment of disease. However,
one must be cautious to acknowledge the fact that it will be easier to de-
monstrate the benefit to society of research in general than of any particular
proposed research project. Thus, historically, one can review research in a
field (eg, development of antibiotics) and show the enormous benefits that have
accrued to society through the development of penicillin, tetracyclines, and
so on. However, some equally well-conceived research projects have yielded no
valuable drug. All research is conducted with an awareness of the a priori
assumption that one does not know in advance what the outcome will be of that
particular experiment. Thus, when speaking of a particular research proposal
one can only discuss potential or hoped for benefit.
Well-done research that proves with certainty that a specific diagnostic
or therapeutic maneuver is not valuable (not safe or not effective) also
benefits society. The benefit, in fact, may be even greater than that de-
riving from "successful" developmental research. It may save society the
cost of further development of the unsafe or ineffective modality. It may
further save individuals the risk of being exposed to the intervention found
to be either unsafe or ineffective.
Benefits of research in this category have major economic and social
implications. Economically, one might consider the amount of money saved
by virtue of development of successful immunizations to polio or rubella.
2-33
From a social point of view one might consider how the development of im-
proved technology changed the social status of individuals who contracted
leprosy or syphillis. One might also consider how the development of psy-
chiatric techniques first changed the social status of individuals once
branded "insane" and subsequently, the economic and social implications
of the development of phenothiazines.
2) Basic research involving human subjects designed to develop im-
proved understandings of the biology and psychology of normal and abnormal
health
(diseased) human beings. Ordinarily, this research offers no direct^ benefit
little
to society or to the individual. However, withoi . this research, research
A
in category 1 is possible. Thus, the derivative benefits are those of
category 1; however, it is a useful exercise to remind ourselves
that the potential benefits of research in category 2 are more remote
than are those described for category 1.
3) Basic research not directly involving human subjects. When one
considers this sort of research in the context of providing a fundamental
basis of knowledge on which research in categories 1 and 2 might be derived,
one usually considers studies done on "animal models" of human diseases;
basic pharmacological studies, basic virology, basic psychology, and so on.
Other studies that are obviously connected include some which do not even
involve intact animals; eg, basic enzymology, studies of cerebral synapto-
somes, and so on. However, it may be appropriate to consider almost all re-
search as potentially providing health related benefits to society as in-
sights are developed to form the basis for research in categories 1 and 2.
Thus, for example, one might trace the tortuous history of research designed
2-34
to develop rocket fuels resulting in the development of drugs (monoamine
oxidase inhibitors) which are now used widely in the treatment of depression.
Similarly, "mustard gas" was developed as a weapon for World War I; studies
on its effects by various types of investigators resulted in the development
of our first effective cancer chemotherapeutic agent. The development of
cortisone and related hormones was expedited enormously on the basis of (admit-
tedly erroneous) reports of spys that the Germans (then our enemies) had
demonstrated that adrenal cortical extracts enhanced greatly the stress tol-
erance and stamina of Luftwaffe pilots. Anthropologists have made multiple
contributions; eg, the description of the use of the Calabar Bean as an "ordeal
poison" in trials for witch craft in some remote tribes in West Africa. This,
in turn, led to the development of physostigmine--our first effective anti-
cholinesterase. Subsequent research in the same area also led to the develop-
ment of the organophosphates, one of which, DFP, not only is effective in
the treatment of glaucoma but also is the widely publicized "nerve-gas".
4) It might be appropriate to sub-divide social research as we have
biomedical and behavioral research. Thus, applied social research might be
discussed as was applied biomedical and behavioral research (performed with
the intent of developing or perfecting diagnostic and therapeutic modalities).
An attempt to develop this analogy was presented in the boundaries paper. Thus,
a social scientist may somehow perceive that there might be some societal dys-
function. His research may be designed to describe or analyze this dysfunction
sufficiently adequately so that he might recommend a "remedy". This remedy
might be actualized by changing public policy or by changing the procedures
within an institution.
2-35
The second category of social research might be analagous to category 2
of biomedical and behavioral research. That is basic research involving
human subjects (done with the intent of improving our understanding of the
structure and function of various societies, institutions, and so on). This
sort of research often forms the basis or provides the research "leads" for
the activities of applied social researchers.
Another benefit to society of social research is that it enriches our
understanding and enjoyment of our selves, our species and our interactions.
It is difficult to assess this either quantitatively or from a utilitarian
point of view. But how many people who will never go to Samoa have enjoyed M.
Mead's books? How many are fascinated by The National Geographic? On the
other hand, when E. Bowen (L. Bohannan) wrote Return to Laughter she could
not have imagined that it would one day be used in the orientation of Peace
Corps volunteers going to Nigeria.
As the Commission has observed earlier: "Scientific inquiry is a dis-
tinctly human endeavor." (Research on the Fetus, p 63). Through publications
of the sort mentioned above, nonscientists may participate vicariously in these
distinctly human endeavors.
Obviously this benefit is not limited to social research.
Benefits to subjects
Direct health benefits
It will be more difficult to describe benefits to subjects than to
describe benefits to society. As mentioned earlier, all research must proceed
withthe a priori assumption that one does not know in advance the outcome. This
becomes a particular problem in discussing with individual subjects the potential
benefits of simple (as defined in the boundaries paper) activities. This is not
2-36
to say that one does know with certainty what the outcome will be--what the
direct therapeutic or diagnostic benefit will be--of a simple activity that
has been classified as standard or accepted. However, by the time a simple
activity has been classified standard or accepted it is ordinarily possible
to inform the patient what the probabilities are of direct therapeutic or
diagnostic benefit and what the probabilities are of adverse effect. It also
is at this point possible to present a reasonably comprehensive list of what
the beneficial or adverse effects might be.
Before discussing this issue further it might be of value to re-examine
the concept of innovative therapy described in the boundaries paper. The
ability to predict diagnostic or therapeutic benefit proceeds along a con-
tinuum. During the earliest stages of working with a innovative diagnostic
or therapeutic technique, ones ability to predict the outcome is very small- -
in many cases it approximates zero. With successive repetitions of the in-
novative intervention the power to predict the outcome increases. By the
time a procedure ceases to be considered innovative- -the point at which it
is considered standard and accepted — the power to predict the outcome of its
implementation is ordinarily nearly as high as it will ever be; it is virtually
never 100 per cent. Clear exceptions to the preceeding statement exist. Thus,
for example, after a modality is considered standard or accepted, continuing
investigation or experience may change our understanding of what the outcome
of its implementation might be. Thus, new toxicities or new benefits might
be discovered.
One may acknowledge the fact that in consideration of complex activities
2-37
(cf boundaries paper) it may be appropriate to tell the prospective subject
that there will be direct diagnostic or therapeutic benefit. However, the
benefit will ordinarily derive from those aspects of the complex activity that
may be considered practice rather than research. For example, if one wishes
to study the effects of chlorothiazide (a diuretic) on sodium balance in patients
with congestive heart failure and if one selects subjects in whom chlorothiazide
is indicated and administers the drug in appropriate doses, the subject may re-
ceive direct therapeutic benefit. This study might be accomplished in a metabolic
research ward. It might involve a period of two or three weeks of eating a
constant diet with precise control of sodium content. It might involve repeated
samplings of venous blood and collection of all urine excreted during those two
or three weeks for purposes of sodium assay. The patient may be expected to
receive direct therapeutic benefit through administration of the drug; however,
this is only technically a research intervention. The subject will not benefit
ordinarily from repeated blood sampling and urine collection. The subject may
or may not benefit from the constant diet; it might be more or less nutritious
and/or palatable than the diet to which he is ordinarily accustomed. The subject
might also benefit from a two or three week period of relative rest on a metabolic
research ward. If the subject required hospitalization for that long a period
of time anyhow it is likely that he will find the accommodations better on the
metabolic research ward than on the usual hospital ward; however, this is not
true in all hospitals. If the subject did not require hospitalization for thera-
peutic purposes, the period of incarceration might be viewed more as an inconvenience
than as a benefit. Further, owing to the customary practices of very careful
scrutiny of all activities on metabolic research wards, the subject might derive
2-38
additional benefits as follows: Any adverse effects of the chlorothiazide
are likely to be discovered earlier than they would in the course of the
ordinary practice of medicine. Thus, the risks of taking the drug would
be reduced accordingly. Further, any complications of the subject's basic
disease are likely to be found and tended to quite promptly. Additional
ramifications may be provided if desired. Also, additional examples may be
provided. However, this illustration provides a general paradigm of the sorts
of direct health benefits that may accrue to a subject of research.
Cohen :
The Hawthorne effect (Levine, and Clinical Research 22: 111-112, 1974)
A
which is well known to social scientists also greatly influences the outcome
of many behavioral and biomedical research activities. Thus, it is often
found that patients in a research environment seem to do better than patients
receiving the same intervention in a practice environment. The probable ex-
planation of this phenomenon is that subjects respond to the enthusiasm and
optimism of the personnel found in the research environment. A dramatic
example of this may be found in the studies of Rashkis and Smarr (Arch. Neurol.
Pschiat. 78: 89-94, 1957).
To this point we have been concerned with the direct benefits that may
accrue to a subject by virtue of his participation in either biomedical or
behavioral research. The benefits may be analyzed similarly whether the re-
search is in the biomedical or in the behavioral field. Subsequent discussion
In this paper will be concerned with benefits categorized as follows: Psy-
chosocial, economic, and derivative. It should be made clear at the outset that
2-39
discussion of these benefits is not meant to be a discussion of the benefits
of either psychosocial or economic research; rather, it is meant to be a
discussion of benefits of any biomedical or behavioral (and, to a limited ex-
tent, social) research; ie, what is being discussed is the category of benefit
and not the category of research. It should further be understood that as one
moves along the spectrum from direct health benefits to the other categories
of benefits in the order listed there will be an increasing tendency to view
these benefits as coercions. Specifically, explicit or implicit offering of
these benefits during the process of negotiating with a prospective subject
to obtain informed consent may be coercive. This paper will not deal further
with the extent to which they ought to be considered coercive or whether use
of such coercions is appropriate.
Psychosocial benefits
Some direct psychological benefits were discussed earlier in consideration
of direct health-related benefits. These might be illustrated more dramatically
if instead of considering the use of chlorothiazide in a complex activity we
used for purposes of illustration of chlorpromazine, shock therapy, or T-groups.
Similarly, the general improvement in morale implicit in the discussion of
the Hawthorne effect may, to some extent be considered a direct psychological
benefit.
The foregoing is essentially a shopping-list of other sorts of psychosocial
benefits any or all of which the Commission might wish to explore in detail.
Patients who know they have terminal illnesses and patients who are depressed
for any reason will often respond favorably to the notion that investigators
are not only interested in them but also attempting to devise something that
2-40
might offer relief or, perhaps, cure. Some patients with cancer in whom all
"accepted" modes of therapy have been tried without success will experience
a psychological "lift" when the option of trying an experimental drug is
offered. In this context the patient (now become subject) may be relieved
to learn that he need not give up hope as there is yet another possibility
for relief (cure?). Further, the subject will often be relieved to learn
that he is not about to be abandoned by the physician (now become investigator).
Many individuals who are either depressed or anxious (or both) will experience
relief as they assume the role of subject; in the relatively sheltered research
environment they are largely divested of the burdens of some sorts of decision-
making.
Individuals who are concerned about their sense of worth may welcome the
opportunity to appear valuable to themselves as well as to others. Often they
may achieve an enhanced sense of personal worth through doing something that
they consider altruistic. Among the obvious examples of such individuals are
prisoners (who might, incidentally, hope that their altruistic tendencies will
be appreciated by those who make parole decisions) and some women who are about
to have abortions.
In some social groups playing the role of subject may bring an individual
considerable prestige. Here we are not only talking of sociologists who might
consent to playing the role of biomedical research subject in order
to complete doctoral dissertations. There are ordinary citizens who are flat-
tered to be the subject (or object) of attention of so many "important" people.
This is particularly true of individuals who become eligible for the role by
virtue of having a rare disease. Other persons eain what ♦'hey consider to be
2-41
substantial prestige or satisfy their tendencies to exhibitionism through
participation in research that attracts great publicity; eg, Reed's studies
on yellow fever, open-heart surgery, kidney transplants (particularly the
role of donor), and so on,
Economic benefits.
The most obvious economic benefits are direct cash payments to the sub-
jects. Some individuals who are filling the roles of both subject and patient
in complex activities may be offered free health care most or all of which
they would have had to pay for themselves had they not agreed to play the role
of subject. Consent to play the role of subject may afford the individual
access to improved accommodations in hospitals, prisons, or other institutions;
to better food and/or medical care or to improved prerogatives of various sorts
when the investigator and subject simultaneously play any of the following com-
binations of role-relationships: doctor-patient, employer-employee, teacher-
student, warden-prisoner, and so on.
Derivative benefits
Let us now consider the benefits of research which brings no personal
health or economic benefit directly to the individual who volunteers to play
the role of subject. Here we will be concerned with what the subject ex-
periences as personal benefit as a function of his sense of kinship with the
individuals who might derive direct benefits. To some extent derivative
benefits intersect with some phenomena described as psychosocial benefits.
To the extent that the subject feels a sense of kinship with larger and larger
groups of humans (the largest group being the entire human species) the moti-
vation increasingly approximates pure altruism. As discussed earlier, a sense
2-42
of altruism may in some cases be experienced as a psychosocial benefit.
Perhaps the closest sense of kinship one might feel is with ones
self. Thus an individual might be motivated to take risks as a subject of
a study (category 2) designed to explore the biology of the disease with
which he is afflicted. The hoped for benefit might be that--based upon the
basic knowledge developed in the course of this research — subsequent re-
search might develop a therapeutic intervention that would provide a direct
health benefit to him . Alternatively, the subject might feel a sense
of kinship with others with the same disease hoping that in the future some
direct benefit might accrue to them. Derivative benefits in this category
are most likely to influence individuals with prolonged chronic diseases.
Thus, for example, an individual with cirrhosis may be motivated to take
substantial risk--perhaps including serial liver biopsies--with the hope that
information might be developed that would lead to the further development of
a therapeutic intervention that might bring direct benefit to him within the
subsequent 2 or 3 years.
Another close kinship relationship would be that of the family. This
is particularly likely to be a motivating factor in consideration of research
in genetics and transplantation . . Thus an individual might be more
willing to offer a kidney to a close relative than to a stranger. The
father or mother of a child with phenylketonuria might be willing to
participate in research designed to perfect techniques for detecting carriers
even though it will bring no direct benefit to them or to their child; they
already know they are carriers. However, if a better method for carrier-de-
2-43
tection might be discovered it might provide direct health benefit to another
relative who is phenotypically normal but who might be a carrier.
Senses of kinship might be based on racial or ethnic factors. Thus,
some Jews might be motivated to serve as normal controls for research de-
signed to explore the pathogenesis or therapy of Tay-Sachs disease; blacks
might volunteer for similar roles in research related to sickle-cell anemia.
Women who are about to have abortions may feel a sense of kinship with
other pregnant women who expect not to terminate their pregnancies but rather
to continue to term; among such women in the future might be the woman who
is now planning an abortion. Thus, she might be motivated to participate in
research made possible by virtue of the fact that she has planned to have an
abortion and designed to develop knowledge that might be of benefit to
pregnant women who expect to carry their pregnancies to term.
The preceding discussion of derivative benefit based upon "senses of
kinship" is related to the arguments raised by Commissioner Lebacqz in which
she invokes the "principle of proximity" (Research on the Fetus ; p-87).
THE " ROLE " OF ASSESSMENT OF RISK - BENEFIT CRITERIA IN THE DETERMINATION
OF THE APPROPRIATENESS OF RESEARCH INVOLVING HUMAN SUBJECTS.
The preceeding pages represent an attempt to incorporate all of the factors
that might be considered in determining the appropriateness of research involving
human subjects as either risks (or in some cases costs) or benefits. Thus, it
is now proposed that, when each of these factors has been considered adequately
and assuming that one will not proceed without adequate informed consent, no
other factors need be considered.
2-44
Ordinarily one would object to such a proposal on grounds that risk-
benefit analyses used for these purposes have a tendency to become purely
teleologic endeavors. Yet, there are incorporated under risks to society
some of the sorts of issues that are commonly used in deontologic analyses.
In particular, some sorts of activities that might erode our concepts of
personhood or our assumptions of trusting relationships with one another
and with various institutions ought to be given very heavy weight in the
determination of the appropriateness of any particular research proposal.
It may well be that one might wish to draw an absolute line as to how far
one might permit these concepts and assumptions to be modified for research
purposes. As was illustrated in the debates leading to the recommendations
on the fetus it will be most difficult to achieve consensus on where the line
ought to be.
Some other factors have been incorporated in the preceeding discussion
that are not necessarily considered by all as risks or benefits. To reiterate
briefly: It was pointed out that research that is badly designed is not
likely to yield any benefit either to the subject or to society. Thus, it
was proposed that putting human subjects at risk in badly designed research
that
is not ethical. This concept was extended as a requirement that research puts
human subjectsat risk should not only be well-designed but also conducted by
investigators who have sufficient skill and adequate facilities not only to
conduct the research but also to protect the rights and welfare of the subjects.
Similarly, it is inappropriate to put humans at risk to gain information that
could be secured without putting humans at risk. Finally, reference was made
to possible mechanisms for compensating individuals who might be harmed by re-
2-45
search.
There remain two additional problems: 1) Should one attempt to quantify
the risks and benefits of any particular research proposal? If so, how? 2)
Who has the authority or the responsibility to analyze the ratio between risk
and benefit of any particular research proposal to determine whether or not it
is appropriate to involve human subjects in it?
Definitions of role and appropriateness
The definitions of the words role and appropriateness are those con-
tained in Webster' s Third New International Dictionary . Since not all of
the definitions contained in that dictionary are germane to the present dis-
cussion, the following is a clarification of which definitions are to be used.
Role , noun: la(l) : A character assigned to :>r assumed by someone; (2):
A socially prescribed pattern of behavior corresponding to an individual's
status in a particular society. b(l): A part played by an actor; (2): A
part assumed by a singer.
2: A function performed by someone or something in a particular situation,
process, or operation (eg, the role of the teacher in the educational process)
(the role of automobiles in leisure has been significant).
Definitions (1) and (2) under la will be important in discussing the
roles played by various individuals and groups of individuals in making de-
cisions on whether a particular research proposal is appropriate. Thus, for
example, the IRB might be assigned the role of determining whether it is ap-
propriate to invite an individual to become a subject of a particular research
project. Similarly, that individual might be assigned the role of determining
whether his participation in that particular project is appropriate (from his
own perspective). In this process, each according to their (his) own criteria,
the IRB and the prospective subject will have assumed or will have been assigned
the role of decision-maker.
Definition b will be used to discuss the roles played by individuals in
the process of conducting research; it will also be used to illustrate thac
2-46
these individuals should be aware that alternate roles are ordinarily available
to them. Thus, for example, an individual may choose to play the role of
subject; alternatively, in some situations he might rather choose to play
the role of patient. Similarly, the professional might have available to him
the alternatives of playing physician or investigator or some combination there-
of.
Definition 2 most closely approximates the meaning ot the word when used
to discuss the role of assessment of risk-benefit criteria. . .and so on. Here
one is concerned with the function performed by something (in this case, as-
sessment of risk-benefit criteria) in a particular process or operation (in this
case, the determination of the appropriateness of research involving human sub-
jects).
Appropriateness , noun: The quality or state of being appropriate.
Appropriate , adjective: 1: Specially suitable: Fit, proper. 2: Be-
longing peculiarly: Special.
In this paper, the word appropriateness will be used as a noun meaning:
The quality or state of being fit or proper.
Parenthetically, it might be noted that the word, appropriate, is also de-
fined in the dictionary as a transitive verb. The etymology of this word is
very similar to that of the adjective. This word will not be used as a verb in
this paper. It should be noted that the verb appropriate has for its synonyms:
Preempt, usurp, arrogate, confiscate, pilfer and purloin. All of these verbs
mean to seize or to take over more or less dictatorially. It should be clearly
understood that none of these meanings are intended in the use of the noun, ap-
propriateness, or the adjective, appropriate, in this paper. The connotations
of the verb are in general the very things one wishes to avoid as one determines
the appropriateness of research involving human subjects.
2-47
Quantification of risks and benefits
Current DHEW regulations (CFR 45 A, Section 46.2 (b) (1)) state the
criteria for determining the appropriateness of research. In particular, it
charges the IRB to: "(b) determine whether. .. subjects will be placed at risk,
and, if risk is involved, whether: (1) the risks to the subject are so out-
weighed by the sum of the benefit to the subject and the importance of the
knowledge to be gained as to warrant a decision to allow the subject to accept
these risks..." (Emphasis added). Additionally, HEW guides for preparing ap-
plications for grants or contracts ( NIH Guide , Vol. 3, No. 12, August 26, 1974)
require that the applicant: "6: Analyze the risk-benefit ratio." Implicit in
these regulations and guidelines is the assumption that the IRB can balance in
some formal or quantitative way the risks to the subject against the benefits
not only to the subject but also--in the clause calling for importance of the
knowledge--to society. The "guidelines" are not explicit as to how risk-benefit
analyses are to be conducted.
One might infer from these regulations and guides that there is necessity
for quantitative analyses of risks and benefits. Supposedly, if the benefits
"outweigh" the risks one should make a decision to proceed with the research.
Proposals have been made to analyze risk (cost) -benefit ratios according to
some sort of formal mathematical equation. In general, what has been suggested,
is that the decision-maker--whether it be an individual or a group—develop
some criteria for assigning to any given predicted outcome a value; this value
should be stated in numerical terms. Thus, risks would be assigned negative
values and benefits, positive values. To derive a "sum" one would add together
all of the values. Since, as has been discussed earlier, risk represents a chance
of harm and most benefits- -particularly benefits contingent upon the "success" of
the experiment --must also be expressed as chance or probability of deriving benefit,
each of the numerical values must be multiplied by some number (something between
2-48
zero and one) which expresses the probability of the risk becoming manifest
as harm or the realization of various uncertain benefits. The various devices
that may be used to attempt to express these analyses in quantitative terms
are covered comprehensively by Wagner, H.M. : Principles of Operations Research ,
Prentice - Hall , Inc., Englewood Cliffs, New Jersey, 1969.
The fact that risks may be given different weights depending upon their
probability on one hand and, on the other, their severity has previously been
clearly expressed by the Commission ( Research on the Fetus , Report and Re -
commendations , pp 83-85). This statement reflects the complexity of analyzing
risks to the individual research subject--in this case, the fetus. It also
reflects the fact that--even in this limited context—there is no way to assign
either a weight or a probability to the risk of one particular harm; viz, the
experience of pain. Of itself, this statement provides suitable cause for not
attempting to determine risk-benefit ratios quantitatively. The difficulties
in developing mathematical models designed to permit decision-making on a
quantitative basis become increasingly complex as we attempt similar analyses
of benefits (or probabilities of hoped for benefits). Attempts to balance risks
to the individual against benefits to society obviously present even more of
a problem.
At this point it seems appropriate to avoid using mathematical models to
calculate risk-benefit ratios for purposes of determining the appropriateness
of research. This statement should not be construed as opposing experimentation
with such mathematical models. Continued attempts to devise such models may not
necessarily result in the development of a suitable (safe and effective) decision-
making model. However, attempts to devise such models often reveal to those
participating in the attempt the information they lack--and, perhaps, should
seek--which would permit "rational" decision-making.
2-49
What constitutes an adequate description of risks and benefits ?
An adequate description of the risks and benefits of any particular re-
search proposal should reflect a consideration of each category listed under
risks and benefits not only to the subject but also to society. Ordinarily,
this description will be provided in a protocol prepared by an investigator
for review by an IRB. This assumes that all investigators proposing research
on human subjects will be aware of all categories of risks and benefits germane
to their particular proposals. This assumption is clearly not correct.
In general, the IRB (as a group) should be aware of each of these cate-
gories. Thus, as they review any particular research proposal, they should
detect any categories that have been omitted and request of the investigator
an adequate description of the risks (or benefits) in the omitted category.
Such interactions between the IRB and investigators have the effect of edu-
cating all concerned (IRB members as well as investigators).
What constitutes an adequate description of any particular risk or benefit?
In the preceeding section it was suggested that each risk and each benefit might
be expressed as a function of its magnitude multiplied by its probability. This
suggestion- -which sufficed for purposes of the preceeding section--is an over-
simplification. A truly adequate description of a risk or a benefit requires
further analysis of both probability and magnitude.
An adequate description of the magnitude of either a harm or a benefit
should include as complete a statement as possible of its expected duration.
Thus, for example, if a possible harm is paralysis of one leg, how long is this
paralysis expected to last? Certainly, the magnitude of the harm will be con-
sidered greater if the paralysis is ordinarily expected to last one year than
if it were expected to last one hour. If the harm is ordinarily expected to
2-50
be irreversible--ie, expected to continue unabated for the duration of
the subject's life--this represents the greatest possible magnitude of that
particular class of harm.
In some situations, potentially irreversible harms, if detected early,
may be either avoided entirely or reduced in magnitude. This avoidance or
reduction may be accomplished by discontinuation of the potentially harmful
research procedure. Alternatively, minimization of a developing or nascent
harm may be accomplished by therapeutic intervention; eg, timely administration
of an antidote to a "poison". In such circumstances a fully adequate description
would include a listing of procedures that might be employed for timely detection
of the developing harm. There should also be a clear statement of what criteria
will be used to determine when to terminate the research or administer the an-
tidote. There should further be an assessment of the probability and magnitude
of success that can be reasonably expected of the monitoring procedures and
corrective interventions.
Similarly, the magnitude of a hoped for benefit should be analyzed in terms
of its expected duration. For example, if the research modality provides the
hoped for benefit, what provisions have been made to assure the subject's con-
tinuing access to this (now proved beneficial) modality? The beneficial modality
might be a drug or a device which proves effective (beneficial) in a particular
subject but whose sponsor (eg, industry) decides to discontinue producing it
because it has not been found beneficial to a sufficient number of individuals
to make its further development worth the sponsor's investment. Alternatively,
it might be an "experimental" health delivery system developed under public
2-51
funding in a community lacking the economic resources for its continuation
at the termination of the period of public funding. For another example,
the compensation for harm caused by research (payment for medical expenses,
rehabilitation, and so on) may terminate (because the period of funding of
the research has expired) before the rehabilitation of a particular harmed
subject has been completed.
Similarly, the probability of the occurrence of both harms and benefits
should ordinarily be elaborated. In consideration of harm, is there any means
by which individuals who are most susceptible to harm might be identified? If
so, will these means be used and will those individuals either be excluded from
the research or informed that they are especially vulnerable? For example, in
planning research designed to test the effects of strenuous exercise in "normal"
humans, one would ordinarily plan to perform various screening tests to identify
individuals with coronary artery disease in order to exclude them.
An even greater problem is presented when it is proposed to recruit re-
search subjects from populations that either have limited capacities to consent,
have subordinate relationships to the investigator or his institution (patients,
students, employees, and so on), or- -by virtue of other aspects of their life
situations—are especially vulnerable (infra). The variety of problems men-
tioned in the preceeding sentence are all to be discussed in detail in subsequent
papers planned by the Commission. At this point, it should be noted that one
of the factors that must be taken into account in determinations of the ap-
propriateness of any proposed research is whether it is proposed to use as
subjects individuals in these categories.
Existing and proposed DHEW regulations have identified as classes of persons
having "limited capacities to consent" children, fetuses, prisoners, and the in-
stutionalized mentally infirm. To this list at least the following should be
added: the unconscious and the inebriated (by alcohol, marijuana, narcotics,
2-52
LSD, and so on). Other persons might be categorized as being especially
vulnerable as a consequence of their life situations. For example, there
are those who are legally enfranchised to give consent but in reality are
incapable of sufficient comprehension to do so (eg, the adult mentally retarded
and psychotics who have not been declared legally incompetent). Poor persons
might be especially motivated to take risks in return for economic benefits.
Persons having prolonged chronic illnesses (or other incapacities) may be
highly motivated to participate as subjects of research that offers any possi-
bility of relief. Further, they may be highly motivated to participate in re-
search designed to develop basic knowledge which might subsequently be used to
plan research designed to develop therapy for their conditions. This becomes
a particular problem when such people perceive themselves as "desperate" and
willing to take "any risk" for even a remote possibility of relief. In this
category are some infertile persons or couples who "desperately" want a child;
some persons with chronic, painful, and disabling disorders such as rheumatoid
arthritis; some obese persons who cannot lose weight following standard pro-
cedures; and so on. Depressed persons and others who question their self-worth
are peculiarly vulnerable (cf, psychosocial benefits). An especially vulnerable
group are some of those who believe (correctly or incorrectly) that their own
death is imminent; some of these people may be highly (and, at times, inap-
propriately) motivated to assume substantial risk without any expectation of
direct health-related benefit.
Those having the authority to determine the appropriateness of research
proposals should scrutinize with particular care those proposals which might
involve subjects with limited capacities to consent, subjects existing in in-
2-53
herently coercible relationships to the investigators, and other subjects who
have been identified as vulnerable. In general, one should be particularly
concerned that the process of recruiting subjects does not even inadvertently
capitalize on their vulnerabilities. The paper on guidelines for selection
of subjects will address these problems in detail.
In consideration of benefit, is there any means by which individuals who
are most likely to be benefited might be identified? If so, will these means
be used to assist in recruiting research subjects who are most likely to be
benefited? A necessary consequence of using such means is the exclusion of
those who are relatively less likely to receive benefit. Thus, in the de-
velopment of a therapeutic innovation—particularly one designed to alleviate
a serious disorder or one whose administration or implementation entails con-
sequential risk- -it is generally most appropriate to select subjects in whom
standard modalities have been tried without success.
Who has the authority or responsibility to assess risk - benefit criteria
in the determination of the appropriateness of research ?
For an extremely thorough survey of this issue the reader is referred
to Katz, J.: Experimentation with Human Beings , Russell Sage Foundation,
New York, 1972.
A central role should be assigned to the IRB in the assessment of risk-
benefit criteria. A more thorough description of the IRB--its structure and
functions- -will be the subject of another paper. The following discussion
will briefly outline how the IRB can discharge this particular responsibility,
how it might analyze risk-benefit criteria differently in relation to different
2-54
sorts of research and subjects, and how the authority of the IRB is, or
ought to be, limited in some cases and, in other cases, extended.
Limitations and extensions of IRB authority
In relation to research supported by DHEW it is now accepted that some
aspects of IRB activity will be subject to review at a national level. Thus,
for example, while the IRB will make determinations regarding the scientific
merit of a proposal, the competency of the investigators, and the facilities
available to them, this review will be repeated by appropriate study sections
at NIH. Similarly, decisions regarding economic priorities--that is which
among all the approved research proposals will be funded- -will similarly be
made at a national level by study sections and advisory councils to the
various institutes.
Some categories of research will also be reviewed at a national level with
the focus of review being on the ethical aspects of the research. In its report
Research on the Fetus , the Commission recommended review by a national
ethical review body for two categories of research on the fetus. In the rules
and regulations promulgated by the Secretary, DHEW, following the receipt of the
Commission's recommendations (August 8, 1975) national ethical review bodies
named Ethical Advisory Boards were established. They are assigned responsibility
for reviewing research on the fetus in the two categories specified by the
Commission in its recommendations. Additionally, they are assigned responsibility
for review of all proposed research involving in vitro fertilization. Thus, it
appears that national ethical review will be focused on those sorts of activities
classified as social risks to society; activities that might erode our concepts
2-55
of personhood or our assumptions of trusting relationships with one another.
Two other types of activity might appropriately be reviewed at a national
level. The first is that sort of activity in which society is the subject and
the risks to society are deemed substantial (in contrast to minimal as defined
earlier by the Commission). The second type of research that one might wish
to have reviewed at a national level is that in which the risks might be borne
by subsequent generations. Thus, one might include in this category research
and development procedures that might lead to pollution (broadly defined) of
the environment or to genetically determined anomalies.
In most institutions having IRBs constructed as prescribed in DHEW rules
and regulations, the IRB is assigned responsibility for review of all research--
not only that conducted or supported by DHEW. It is also clear that most
funding agencies now require "institutional endorsement" indicating that a
research proposal has been reviewed by the IRB; these agencies include private
foundations, drug industry, and others. It seems safe to assert that these
agencies, like DHEW, have mechanisms analagous to those of DHEW for making
determinations of scientific merit and economic priority.
Role of the IRB
With regard to assessment of risk-benefit criteria it is the responsibility
of the IRB to review research protocols prepared by investigators. The IRB
should first determine whether there are, in fact, any risks to the subjects
or to society. If there are no material risks there is probably no need for
further review.
If the IRB determines that there are risks it should proceed as follows:
It should determine that the investigator has designed the protocol so as to
minimize risk and maximize the chance of benefit. It should then attempt to
determine whether: "The risks to the subject are so outweighed by the sum of
the benefit to the subject and the importance of the knowledge to be gained
2-56
as to warrant a decision to allow the subject to accept these risks." The
assumptionscontained in this statement will be discussed in some detail in
the next section.
In making this determination, the IRB will be obliged to consider the
characteristics of the proposed subject population. In particular, is it
proposed to recruit research subjects from populations that either have
limited capacities to consent, have subordinate (inherently coercible) re-
lationships to the investigator or his institution, or--for various other
reasons- -have been identified as especially vulnerable ( supra )? In general,
when such subjects are to be used there will be a requirement for lower risk-
benefit ratios than in research proposing to ase other sorts of subjects.
Further, there will, in general, be a requirement that the hoped for benefits
be, to the extent possible, direct health related benefits (to the subject)
rather than other categories of benefits; in these subjects benefits that have
been categorized as psychosocial, economic, and derivative, might be seen more
as potential coercions than as hoped for benefits. Finally, IRB review should
determine whether the hoped for benefit (knowledge) might be obtained if the
research were conducted in less limited, coercible, or otherwise vulnerable
populations; if so, use of vulnerable subjects should be avoided (cf, next
section) .
Having made the determination called for in the above cited quotation from
DHEW regulations, the IRB should next review the plans for the consent negot-
iation. With regard to risks and benefits these will ordinarily be described
on a consent form. The IRB should determine that there is an adequate des-
cription of risks and benefits (as broadly defined in this paper) to permit
the prospective consentor to make a rational decision.
The role of the subject
Are we to consider the role of the subject to be a right to which all
2-57
citizens are entitled unless they are deprived of this right by due process
of law? Alternatively, are we to consider it a responsibility which all
citizens must assume when called upon? Alternatively, we might consider it
neither; it might, for example, be considered a job for which one must be
compensated appropriately. In the latter case appropriate compensation
might be any of those compensations described in
the section on benefits to the individual subject.
In considering this issue we may assume that agreement exists that it
is in the interest of society to conduct research designed to improve the
health, education, and welfare of its members and that most of this research
cannot be conducted without human subjects. Thus, it is in the interest of
society to have individuals who will assume the role of experimental subject.
Most roles that are valued by society are seen as either rights, res-
ponsibilities, or jobs. Obviously, these categories are not exclusive. In
fact, most roles that we value are seen as jobs which any individual has the
right to perform provided he has the qualifications and has not been deprived
of that right by due process of law. Depending upon the outcome of these
deliberations we may find outselves presented with some novel additional
questions. For example, as we consider developing guidelines for protection
of individuals who--in our view--have limited capacities to consent--we must
also consider the possibility that we are depriving such individuals of what
they might consider their rights. To the extent we consider the role of sub-
ject as a job, we should decrease our concern with what sorts of benefits or
rewards are permissible. Perhaps we should consider allowing the rewards to
be determined- -to some extent--by customary market factors.
In the framework of our current societal practices and customs it is re-
latively easy to view the role of subject as a right which in some cases may
also be a job. The possibility that in some circumstances it might be ad-
2-58
vantageous to consider it a duty--as we have considered military service--
may be less apparent. This possibility and its implications—particularly
in relation to research on children—has been discussed in detail in a
provocative paper by Alexander Capron (Clinical Research . 21: 141-150,
1973)
Another fundamental question that must be asked can, perhaps, be
disassociated from the analysis of the role of the research subject generally.
Can an individual who is fully (materially) informed and reasonably free to
make choices decide- -for whatever reasons he considers appropriate- -to
participate in research that offers little or no direct benefit to him
even though there is a considerable amount of personalrisk? If the answer
to this question is yes, the role of the IRB in relation to risk-benefit
analysis would be limited to assuring that there is appropriate disclosure
of risks and benefits and so on. In other words, in dealing with such pros-
pective research subjects, the IRB would not be empowered "...to allow (or
disallow) the subject to accept these risks". A particularly provocative
consideration of the ramifications of a positive answer to this question
is presented as The Case of the Kamikaze Astronauts (Katz, pp. 175-176).
2-59
THE NATURE AND DEFINITION OF INFORMED CONSENT
IN VARIOUS RESEARCH SETTINGS
Robert J. Levine, M.D.
December 1 , 1975
The Commission is charged with the responsibility to consider."
"The nature and definition of informed consent in various research
settings." This paper is an attempt to identify and to begin to analyze
the issues that must be considered as one attempts to respond to this
charge. The following assumptions limit the scope of this paper: It is
assumed that the reader will be familiar with the concepts developed in
two previous papers prepared by the author for the Commission (1,2); this
paper contains many references to the two previous papers. Separate papers
are being developed for the Commission on research involving children,
prisoners, the institutionalized mentally infirm and psychosurgery. Since
it is assumed that these papers will deal with the nature and definition
of informed consent in these sorts of subjects having limited capacities
to consent, this paper will not address problems peculiar to those popu-
lations. Further, a separate paper: "...shall consider the appropriate-
ness of applying (these) principles and guidelines. . .to the delivery of
health services to patients under programs conducted or supported by the
Secretary."
The paper begins with an identification of the various functions in-
formed consent is meant to serve. The remainder of the paper is organized
in two large sections. In the first there is an attempt to define informed
consent and to analyze its various component processes. Barriers to the
achievement of each of the component processes are also discussed. The
objective of this part of the paper is to provide a definition of informed
consent sufficiently comprehensive to cover almost all contingencies that
might be encountered in negotiating for informed consent to various types
3-1
of biomedical, behavioral, and social research and with various types of
subjects excluding those identified specifically above.
The next major section of the paper is concerned with various pro-
cedures that might be used to assure that informed consent has been
achieved and, further, that it continues to be maintained during the
conduct of the research activity. There is also a discussion of the
various procedures, devices, and personnel that might be used to document
the existence of informed consent and some conditions under which docu-
mentation of informed consent might be either unnecessary or detrimental
to the interests of all concerned.
The functions of informed consent
Katz and Capron have enumerated several functions of informed con-
sent (3 at pp. 82 et seq) to therapy (innovative as well as standard and
accepted) for catastrophic diseases. However, the authors contend that--
at least in catastrophic diseases—informed consent need not necessarily
differ qualitatively as a function of whether the proposed therapy is in-
novative or accepted. Further, the functions are similar to those of in-
formed consent to research identified earlier by the same authors (4).
The functions of informed consent are (3) :
"a) To promote individual autonomy."
"b) To protect the patient-subject's status as a human being." On
superficial examination it might appear that this category does not differ
substantially from function "a". However, Katz and Capron see this as going
far beyond the requirement for autonomy. A prospective subject may exercise
3-2
his autonomy by either agreeing or refusing to agree to assume the role
of subject. If one would protect a subject's status as a human being
one assumes the responsibility to attempt to create a true "joint inter-
prise" or partnership between the physician-investigator on one hand and,
on the other, the patient-subject.
"c) To avoid fraud and duress."
"d) To encourage self-scrutiny by the physic ian- invest igator ."
"e) To encourage rational decisionmaking."
"f) To involve the public." The meaning of this particular purpose
or function may not seem apparent without some elaboration.
"Primarily, the obtaining of consent can be important for the public
relations of a physician or a medical center. The reverse is certainly
true: A physician who develops the reputation of using his patients as
guinea pigs for his studies or medical innovations without their informed
consent will be avoided by those who know that reputation.
"Informed consent may also function beyond the area of public re-
putation and serve to increase society's awareness about human research."
In the preceding sentence, Katz and Capron are addressing particularly
the practice of educating the public through mass media about all aspects
of research involving human subjects. As they see it, part of the moti-
vation for such informational campaigns might be to recruit individual sub-
jects. However, it also results in making the general public a more in-
formed decision maker in several respects. This function of informed con-
sent will not be addressed further in this paper.
Another function of informed consent mentioned in this book but not
designated as a separate category is identified here as :
g) To reduce the civil and/or criminal liability of the investigator
3-3
and his institution. Since a separate paper on the legal aspects of
informed consent is being prepared for the Commission, this paper will
deal with this function of informed consent only to a very limited ex-
tent.
Definitions of key words
The definitions of the key words to be used in this paper are those
contained in Webster' s Third New International Dictionary . Since not all
of the definitions contained in that dictionary are germane to the present
discussion, the following is a clarification of which definitions are to
be used.
Informed , adjective: "1 : Having information: based on possession
of information."
The second definition of the adjective, informed, equates it with
such words as educated, intelligent, and cultivated; thus, examples are
given such as cultivated taste and educated opinion.
The first definition is the one to be used in this paper. The second
definition develops the nuances of what one should be striving for as one
attempts to create a condition that may be defined as informed consent. It
is not merely having some information; it is striving for a situation in
which the prospective consentor may be said to be educated or cultivated
with respect to the proposition to which he might consent.
Information, noun: "Id: the communication or reception of knowledge
or intelligence;" "2b: knowledge of a particular event or situation (synonyms
intelligence, news, advices);" "5: the process by which the form of an object
of knowledge is impressed upon the apprehending mind so as to bring about the
state of knowing."
Inform , verb transitive: "6: to communicate knowledge to: make ac-
3-4
quainted (synonyms: tell, advise, enlighten)."
Inform will also be used as an intransitive verb: "1: to give in-
formation: impart knowledge (synonyms: acquaint, apprise, advise, notify,
advertise)." "These verbs signify to make aware or cognizant (of something).
Inform implies the imparting of knowledge, especially of facts or events
necessary to the understanding of a pertinent matter."
The various verbs listed as synonyms each have importantly different
connotations. The distinctions are developed well in the cited dictionary.
Since each has connotations different from those intended for the verb in-
form as used in this paper, none will be used. Further, the identification
of these five verbs as synonyms has its roots in other definitions of the
verb, inform, which are contained in the dictionary but not repeated in this
section.
Consent : This word will be used both as a verb and as a noun. The
full dimensions of the meaning of this word may be illustrated by pointing
out that- -whether used as a verb or a noun- -it is derived from the Latin
verb, consentire , meaning to feel together as well as to agree. Thus, among
the now archaic or obsolete definitions of the word are: as a verb, to be
in harmony or concord especially in opinion, statement, or sentiment. And
as a noun, the being of one mind.
The definitions of consent to be used in this paper are as follows :
Noun, "la: compliance or approval especially of what is done or pro-
posed by another (synonyms, acquiescence, permission)" "b: capable, deli-
berate, and voluntary agreement to or concurrence in some act or purpose
implying physical and mental power and free action— distinguished from assent ."
"3: agreement among persons usually as to a course of action or concerning a
particular point of view or opinion."
Intransitive verb. "2: to express a willingness (as to accept a pro-
position or carry out a particular action): give assent or approval (synonym:
3-5
agree)."
Assent, noun: "lc : concurrence with approval (sanction)"; "3: agree-
ment with a statement or proposal especially in a matter of minor importance
or one detached from personal concern: mere acquiescence- - distinguished
from consent ."
Assent , intransitive verb: "1: to give or express one's concurrence,
acquiescence, or compliance (synonym: consent)." Among the various synonyms
given for the verb, assent, are consent, accede, acquiesce, agree, and sub-
scribe. "Assent indicates a concurring, either a positive agreeing or more
passive conceding, without expressed doubts or objections. Consent indicates
a complying, granting, or yielding, willing or reluctant, to request or de-
mand. "
It is of interest that the verbs assent and consent may be used as synon-
yms; assent seems to have connotations more of acquiescence; consent, on the
other hand, seems to be a more positive action more closely approximating
subscription than acquiescence. The differences in the nuances of the two
words are developed more thoroughly when they are used as nouns. Thus, one
definition of each is dependent upon its distinction from the other. Con-
sent is much more a manifestation of the will of the consentor; it is much
more an act of endorsement of an agreement. Assent, on the other hand, may
be a manifestation of a relative lack of interest.
Definitions of informed consent
According to Katz (4 at p. 523): "...the concept of informed consent
has been accepted in case and commentary as a cardinal principle for judging
the propriety of research with human beings. Yet law has neither defined
sufficiently well the substance and ambit of informed consent in therapeutic
settings nor determined clearly its functional relevance for human experi-
mentation. Thus, in invoking informed consent like a talisman, lawyers, in-
vestigators, and courts often seem to overlook the fact that it lacks specific
construction and remains an ill-defined concept."
Further (4 at p. 521): "Belief in the idea of individual freedom is a
cornerstone of the Western concept of man and society. The common law nur-
tures and protects individual freedom through the doctrine of self-determin-
3-6
ation, which confers on each person the right to pursue his own ends in
his own way so long as he does not interfere with specified rights of
other individuals or of the community. The requirement of consent is
the primary means for implementing the abstract notion of self-deter-
mination. Tort law, for example, guards a man's property and person
against interferences to which he has not consented. Similarly, a con-
tract comes into being when two or more persons agree with each other
that certain terms should govern their relationship.
"In most commercial transactions, each party is responsible for in-
forming himself about the terms and implications of the contract. However,
when professionals intervene in the lives of others, a higher standard is
imposed upon them. They may be held responsible not only for obtaining the
layman's consent, but also for informing him of the consequences of their
agreement."
Thus, it may be concluded that informed consent is a type of contract
in which one of the contractors (the investigator)--as in all fiduciary re-
lationships—is held accountable for higher standards of responsible con-
duct than are most individuals in creating commercial contracts. Further,
it is clear that the precise definition and nature of informed consent have
not yet been established.
DHEW rules and regulations (5) provide a definition of informed con-
sent which includes a specification of its 6 basic elements:
"46.3(c) "Informed consent" means the knowing consent of an individual
or his legally authorized representative, so situated as to be able to ex-
ercise free power of choice without undue inducement or any element of force,
fraud, deceit, duress, or other form of constraint or coercion. The basic
elements of information necessary to such consent include:
(1) A fair explanation of the procedures to be followed, and their
purposes, including identification of any procedures which are experimental;
(2) a description of any attendant discomforts and risks reasonably
to be expected;
(3) a description of any benefits reasonably to be expected;
(4) a disclosure of any appropriate alternative procedures that might
be advantageous for the subject;;
3-7
(5) an offer to answer any inquiries concerning the procedures; and
(6) an instruction that the person is free to withdraw his consent
and to discontinue participation in the project or activity at any time
without prejudice to the subject."
Components of informed consent
The process of creating a condition that may be called informed con-
sent is commonly seen simplistically as one having two components. The
first component is that of informing- -the transmission of information
from the investigator to the prospective subject. The second component
is that of consenting; this is signified by a declaration on the part of
the prospective subject that he has assimilated the information and that
he is willing to assume the role of subject. Even more simplistically, it
may be assumed that these two components are accomplished sequentially;
ie, after the information is transmitted the individual consents or re-
fuses to consent, DHEW regulations recognize that there may be some dis-
ruption in the sequence of these processes by requiring "(5) an offer to
answer any inquiries concerning the procedures...". Katz (6 at p. 786)
proposes an entirely different way of viewing the process. He sees the
process as essentially an invitation offered by the investigator to the
prospective subject to join him as a partner in a collaborative venture:
"Informed consent- -would entail, if it is truly seen as an invitation,
asking for consent, seeking authorization to proceed, and not making a demand
under the guise of a symbolic egalitarian gesture. It would necessitate
sharing knowledge and admitting ignorance, answering questions and identifying
unanswerable questions, appreciating doubts and respecting fears... It requires
that the interaction between investigator and subject become a partnership,
giving the subject the right to determine what should be done for and with
him, and forcing the investigator to be explicit in what he wants to do and
why. Thus the controversy over the subject's capacity and incapacity to under-
stand, on which the debate about informed consent has focused, is a displace-
3-8
ment from the real issue, which is the dread of an open and searching
dialogue between the investigator and his subject. This displacement
is caused by the unacknowledged anxiety over making the invitation in
the first place."
As individuals communicate with each other toward the goal of achieving
a condition (a special type of contract) designated informed consent, each
of the component processes may occur in any order. It is proposed that the
most appropriate single word that may be used to refer to the total communi-
cations is negotiation. Thus the total process will often be referred to
as negotiating informed consent. Many documents use different words to
name this process. Thus investigators are often admonished or advised to
secure or to obtain informed consent. Such words as secure or obtain do
not capture the full dimensions of the desired interactions; viz, an in-
teraction involving dialogue, encounter and so on. For purposes of this
discussion, the negotiations will be discussed as having four separate
component parts. 1) Informing; 2) assessment of the prospective con-
sentor's comprehension; 3) assessment of the prospective consentor's
autonomy; and 4) consent. Negotiations for informed consent in the real
world never are conducted as four separate component processes.
Informing
In this section the types of information that should be communicated
to the prospective consentor will be identified. There will be eleven
essential elements of information, six of which correspond to those identi-
fied in DREW regulations.
1) There should be a clear statement of the overall purpose of the
research. When appropriate, it should be stated that there is not only
3-9
an immediate purpose but also a larger ultimate purpose. Thus, for ex-
ample, it might be stated that the immediate purpose of this research
is to develop a more sophisticated understanding of normal kidney function.
If the immediate purpose is achieved, one hopes that this information might
contribute to our ability to identify and treat persons with diseases of
the kidney.
One of the most important consequences of stating the purpose of the
research is that it might alert the prospective subject to decline parti-
cipation in research the goals of which he does not share. Thus, for ex-
ample, some individuals might not wish to contribute to the general fund
of information that would enhance our capacities for genetic modelling.
Some others may decline to participate in research that might identify
their racial or ethnic group as having certain qualities; eg, as having
lower intelligence than the general population.
This element of information may partially duplicate number 6 (infra).
In some situations it might be appropriate to not reveal the true purpose
of the research (number 11) .
2) There should be a clear invitation (not a request or a demand)
to the individual to become a research subject. The implications of playing
the role of subject--as opposed to any alternative role available to the
individual-- should be clearly indicated. Most importantly, when one agrees
to play the role of subject, one ordinarily agrees to becorae--at least to
some extent--a means to an end (1 at p. 3).
If the research is of the sort classified as innovative therapy, the
physician- investigator may be held liable for failure to obtain informed
consent merely by virtue of having failed to explain that the procedure
3-10
used represented a departure from customary practice (7 at p. 253).
3) The prospective subject should be informed as to why he has
been selected for participation in the research. Ordinarily this is
because of some specific disease (or other life situation) that he or
one of his relatives might have. In other cases it might be that the
investigator presumes that he does not have that disease or condition;
it might be that he is being asked to serve as a control in studies de-
signed to explore that condition.
In some cases, the first approach to a prospective research subject
might involve some sorts of testing to determine if he is in fact eligible
to be a subject of the proposed research. The performance of diagnostic
or other testing solely for purposes of identifying an experimental subject
population is itself a form of research. In these situations it is essential
to negotiate informed consent to undertake the "pre-screening" procedures.
This negotiation should of itself contain all eleven elements of informing
that are appropriate; in addition, the consequences of being found eligible
for participation in the research project should be made clear. Thus, as one
is negotiating informed consent to participate in pre-screening tests, one
should also inform the prospective subject as to what research he will be
invited to participate in if he "passes" the pre-screening tests. Some of
the consequences peculiar to "failure to pass pre-screening tests" have
been detailed earlier (2 at p. 9).
4) There should be: "A fair explanation of the procedures to be
followed, and their purposes, including an identification of any procedures
3-11
which are experimental." (5, Section 46.3(c)(1). A fair explanation
would include identification of all procedures and interactions that
would be of material interest to the prospective subject. Thus, if
it is proposed to draw a small amount of blood from a vein for purposes
of assaying some chemical, the prospective subject might be expected
to be more interested in knowing how much blood will be drawn, how
often it might be repeated, where he might have to go to have the
procedure done, and what practical consequences to him there might be
of the results of the assay than he would in the details of the assay
technique. Further, he might be interested in who (if not the in-
dividual negotiating the consent) might draw the blood and what his
experience and qualifications are. The language necessary to convey
the meaning of each of these bits of information will obviously vary
enormously depending upon the experience of the prospective subject
with previous drawings of blood. Thus, for example, it might be possible
to relate the amount of blood to be drawn in terms of what fraction it
is of the amount removed when one donates a pint of blood or in relation
to the amount the individual has had drawn for various diagnostic tests
in the past.
Thus, in addition to describing each of the experimental procedures--
particularly as they affect the prospective subject — it is ordinarily ad-
visable to anticipate that the prospective subject will want to know: a)
With whom shall I interact? b) Where will the research be done? c) When
will the research be done? d) How often will the various procedures be
3-12
performed? e) How much of my time will be involved?
a) Most prospective subjects will be reassured to learn that the
individual negotiating for informed consent with him will play a key
role in the actual conduct of the research (cf, who shall negotiate
with the prospective subject?). However, many types of research acti-
vities require interaction with a large number and variety of profes-
sionals and their assistants. In general, it is better to advise the
prospective subject of the numbers and types of individuals with whom
he will interact rather than to surprise him during the course of the
research. Thus, for example, some prospective subjects may have strong
biases against physical examinations by individuals of the opposite sex
or by students.
b) For various reasons prospective subjects will be interested
in where the research is to be conducted. Thus, for example, some
might feel reassured to learn that a questionnaire will be administered
in their own homes; others might regard this as an unwelcome intrusion.
In some cases--eg, in some hospitals--the research unit might be more or
less attractive to the prospective subject than the alternative facilities
he might have to use should he choose the role of patient. A statement of
where the research is to be done will also allow the prospective subject
to assess the amount of inconvenience there might be in traveling to and
from that location.
c) An explanation of when the research is to be done will allow the
prospective subject to determine whether there are any essential time con-
flicts with his own schedule. Some sorts of research are dependent upon
repeating observations at precisely timed intervals. If this is discussed
3-13
frankly at the outset it may be possible to negotiate a mutually satis-
factory schedule; alternatively, it may be found that a subject must
drop out during the course of the research owing to a prior commitment.
d) A precise statement as to how often various procedures will
be performed will also assist the subject in assessing the totality of
his personal commitment of time and other inconvenience. In some re-
search, it is necessary to have various follow-up procedures done at
intervals as long as a year or more. If the prospective subject knows
he will not be available that much later- -because, perhaps, he might
be leaving the country- -he can advise the investigator that his full
participation will be impossible.
e) In explaining how long the research will take there should be
an estimate not only of how much time each component of the research may
reasonably be expected to occupy but also of the total duration of the
research.
In complicated research activities it is occasionally of value to
invite the prospective subject to visit the site of the proposed research
(eg, the metabolic research unit, the office of the investigator, the
physiology laboratory) where he might see the personnel, facilities,
apparatus, and so on, that will be involved. Many types of biomedical,
behavioral, and social research are conducted as components of what has
been identified previously as complex activities (1 at p. 9). In these
cases special care should be given to explaining which of the activities
are done exclusively for research purposes. In these situations, the
bits of information identified as a) through e) should be elaborated
3-14
as follows: If you agree to participate in this research you will be
interacting with (specify) additional types of individuals; it will
take (specify) additional time; procedures that might have been done
n times will be repeated n plus x times; the location will change in a
specified way; and so on.
5) There should be: "A description of any attendant discomforts
and risks reasonably to be expected" (5, Section 46.3(c)(2). This
quotation from DHEW regulations requires some elaboration. Under the
rubric of discomforts one should include not only physical and psycho-
logical discomforts but also personal inconveniences. The sorts of
personal inconveniences that might be expected to occur as a consequence
of participating in research were discussed in the preceding section. The
prospective subject should be advised that many research procedures might
produce a variety of physical or psychological discomforts. Thus, for
example, a naive subject might assume that a spinal tap would produce
some pain by virtue of a needle being inserted low in his back. On the
other hand, it may not be assumed that he will understand that there is
a high probability of a headache which might be quite severe following
the procedure. A person who is being asked to have skin tests for certain
allergies should be advised that if the tests are positive they might itch,
produce transient discoloration or other disfigurement, and so on. A
person being asked to complete a questionnaire should be advised, if ap-
propriate, that he is likely to find some of the questions either irritating
or embarrassing.
3-15
The risks of participation in research have been categorized in an
earlier paper (2). The various physical, psychological, social, legal,
and economic risks to the subject should be identified during the process
of informing. An adequate description of each risk has also been identified
in the same paper; to the extent possible (or necessary) each risk should
be expressed clearly in terms of its probability and its magnitude (2 at
p. 49). A clear statement should be made as to what steps will be taken
to assure the early detection, minimization, and/or correction of various
harms. There should be clear statements as to how the prospective subject
will (or will not) be compensated for whatever harms occur. Legal barriers
(if any) to preservation of confidentiality should be discussed candidly.
In some cases the prospective subject will be called upon to assume
responsibility for minimizing the chance of harm. He will be asked to
perform certain functions during the course of the research to accomplish
this objective. For example, when a woman of childbearing age participates
in a research activity in which there is known or unknown risk to the fetus,
she should be advised that if she wishes to be a subject she should avoid
becoming pregnant. Her plans for avoiding conception should be reviewed
during the consent negotiations. At times, if her plans seem inadequate,
it will be necessary to either exclude her from the research or to ask her
to agree to more certain plans for contraception. She should further be
advised that if, during the course of the research, she deviates from the
plans discussed at the outset, she should advise the investigator im-
mediately.
Another common problem is presented in negotiations for informed con-
3-16
sent to "double-blind" drug trials.
It must be made clear that neither the investigator nor the subject
will know what drugs the subject is taking. This may present a problem
if the subject becomes ill and requires emergency treatment. The
physician who is called upon to administer treatment in an emergency
may find that he is unable to plan such therapy rationally without knowing
exactly what drugs the subject (now become patient) is taking. Thus, the
subject should be advised that he will be provided with a card indicating
that he is a subject of a "double-blind" drug trial; the card should also
include the subject's code number in the trial. Further, on this card
there should be a telephone number at which there will be someone on
24 hour call who has access to the codes for the study. Thus, in the
event that a physician must know immediately what drugs the subject is
taking, he will be assured immediate access to this information at any
time. The subject should be advised of the importance of keeping this
card with him at all times.
Must all risks be fully disclosed? If not, how does one determine
which must?
There are distinct perils to the process of informed consent presented
by overdisclosure and by underdisclosure. Some of the perils of over-
disclosure were documented by Epstein and Lasagna (8). They presented
consent forms of various lengths and thoroughnesses to prospective sub-
jects of a drug study. They found that the more detail was included the
more likely were the prospective subjects to be either confused or intimi-
3-17
dated. In their study they found a remarkably high incidence of refusal
to take an experimental drug based upon its apparent danger. At the
conclusion of the study they informed the individuals who refused that
the drug they were describing- -the drug they refused to accept money
for taking- -was aspirin; many of those who refused were regular users
of aspirin. Almost all of them reported that although they had de-
clined participation in the "study", they intended to continue to use
aspirin essentially as they had before.
In the medical practice context, a study was done to determine
the influence of full disclosure on the willingness of patients to
consent to angiography (9). The publication of the results of this
study includes the consent forms and questionnaires used. The consent
forms seem to contain an adequate description of the risks of the pro-
cedure (a small probability of serious complications which might in-
clude death). In understanding the implications of this study it is
important to know that, since angiography is a diagnostic procedure, its
performance may or may not result in any information upon which further
therapy might be recommended. Of 232 patients asked to consent, all but
2 per cent did. Response to the questionnaire indicated that the majority
of patients were pleased to have the information conveyed in the consent
form. The author concludes that he is convinced of the value of dis-
closure to the extent contained on their form.
Under informing, on the other hand, may be perilous to all partici-
pants in the research process. Inadequately informed subjects may make
3-18
wrong choices. Harmed subjects who had not been informed of the possi-
bility of that particular harm might quite appropriately litigate
against the investigator.
Perhaps the ideal for which one should strive is a "materially"
informed subject. One definition of material risk is provided in a
decision of the District of Columbia Court of Appeals in 1972 (10).
In this case the court held that the disclosure required was determined
by the "patient's right of self-decision" and further that the patient's
right to make his own decision can be termed "effectively exercised only
if the patient possesses enough information to enable an intelligent
choice. "...A risk is thus material when a reasonable person, in what
the physician knows or should know to be the patient's position, would
be likely to attach significance to the risks or cluster of risks in
deciding whether or not to forego the proposed therapy." This definition
might be applied to determining materiality of risk in the research con-
text by changing the words patient, physician and therapy, to subject,
investigator, and research, respectively.
One problem with this definition of materiality is that it puts the
particular physician or investigator in the precarious position of having
to know in advance what harms a particular patient or subject might con-
sider material after they occur. As the court noted in Cobbs v. Grant
(11): "Since at the time of trial the uncommunicated hazard has materi-
alized, it would be surprising if the patient-plaintiff did not claia
that had he been informed of the dangers he would have declined treatment.
3-19
Subjectively he may believe so with the 20-20 vision of hindsight, but we
doubt that justice will be served by placing the physician in jeopardy of
the patient' 8 bitterness and disillusionment". Thus, some courts have
held that in determining issues of materiality of risk, one should adhere
to the reasonable person or "prudent-patient test." How does one determine
what the reasonable person or "prudent prospective subject" might wish to
know? In the context of research such decisions are largely made by the
IRB. The IRB reviews a research protocol and determines the minimum
standards for disclosure of risk in the consent negotiation with prospective
subjects. However, the capability of the IRB to perform this function has
been challenged by some commentators. In particular it is suggested that
since the IRB is dominated by scientists it does not truly reflect the
needs of the reasonable lay-person to know.
Various remedies for this situation have been proposed. Two of the
most extreme are the following:
Hauck (12) has proposed that a "consent jury" be appointed- -much like
a trial jury--from the ranks of non-scientists. He further proposes that
all the risks of each experiment be listed by those having the scientific
expertise to do so--perhaps an IRB. Then, using an adversary system, the
materiality of each risk would be debated before the jury. Each risk would
have an expert advocate for its disclosure and an expert advocate for its
non-disclosure. At the conclusion of the debate the jury would determine
which risks should be disclosed and in what fashion. Two faults of this
procedure should be mentioned. Firstly, like any other system, it assumes
3-20
an average subject; it permits no flexibility on the part of the individual
who is negotiating informed consent to provide more or less disclosure de-
pendent upon the needs of a particular prospective subject. Secondly,
this process would consume enormous amounts of time and energy.
Norman Fost (23) has proposed that a "surrogate system" might be used
to achieve the same sorts of purposes (as well as others) Hauck proposes
for the "consent jury". Fost suggests that the surrogates be selected
from a population that matches as closely as possible that from which pros-
pective subjects might be drawn in all respects but one: They should be
aware that- -for some reason- -they are not eligible to become subjects (al-
though they are asked to pretend that they are). His proposal differs from
Hauck' s in that the surrogates would not meet as a group; rather they would
meet as individuals with the investigator. The surrogate system is de-
signed to inform the negotiator for informed consent of the range and
diversity of factors of material interest to individuals similar to pros-
pective subjects. Fost emphasizes that he is not proposing this system as
a necessary adjunct to the approval of research proposals (emphasis in the
original).
A second extreme remedy has been proposed by DHEW (13, 14). They pro-
pose that Consent Committees should be established to monitor the actual
process of negotiating informed consent with persons having "limited capa-
cities to consent". This process permits flexibility in that it permits
responsiveness to the needs of a particular prospective subject. However,
it shares with Hauck 1 s proposal the fact that it would be enormously cumber-
some (cf, Consent Committees).
3-21
6) There should be: "A description of any benefits reasonably to
be expected," (5, Section 46.3(c)(3)). The types of benefits that might
be described have been categorized earlier as: a) Direct health benefits;
b) Psychosocial benefits; c) Economic benefits; and d) Derivative bene-
fits (2 at p. 29). In most consent negotiations there should be--when ap-
propriate—a thorough disclosure of categories a) and c). Whether or not
it is appropriate to mention categories b) and d) and the extent to which
they might be emphasized will often be controversial. Similarly, the
emphasis that ought to be given to societal benefits in the consent nego-
tiations will also be controversial. In general, these controversies should
be resolved by the IRB. The main determination they must make will be based
on their judgment of the effects of mentioning such benefits to the specific
proposed research subject population. They must determine the extent to
which that specific population might be unduly coerced by the offering- -
implicit or explicit--of such benefits.
a) Direct health benefits. Hoped for benefits that might be discussed
quite analogously to direct health benefits would include direct educational
benefits, direct welfare benefits, improved working conditions, and so on.
The optimal description of a hoped for benefit has been detailed in an
earlier paper (2 at p. 49). It should be emphasized that it is a hoped
for benefit- -not something that can be guaranteed. Particularly in the
category of innovative therapy (innovative practices) it should be emphasi-
zed that, while a major purpose is to attempt to bring direct benefit to
the subject, an additional purpose is to try to develop a systematic body
3-22
of new knowledge. Thus, the prospective subject is not the only in-
tended beneficiary of the activities. Additionally, as a consequence
of this dual purpose, it will ordinarily be necessary to do additional
procedures- -more testing and more monitoring- -than would be necessary
were the individual playing exclusively the role of patient, student,
and so on. As discussed earlier (2), to the extent possible, each
hoped for benefit should be assessed in terms of its probability and
potential magnitude (including, when appropriate, its expected duration).
c) There should be a clear statement of any material inducements
that are offered to the prospective subject in exchange for his parti-
cipation. Such inducements might include cash, food, clothing, im-
proved accommodations, and so on. If the material reward is contingent
upon completion of the project--that is, if the subject will lose all
or part of the reward as a consequence of withdrawal from the project--
this should be specified.
7) There should be: "A disclosure of any appropriate alternative
procedures that might be advantageous for the (prospective) subject"
(5, Section 46.3(c)(4)). The alternatives in general to playing the role
of subject will be discussed in detail below (number 10). In this section
there will be a more narrow focus on research activities in which the
hoped for outcome is to bring a direct health related benefit to the
prospective subject. To some extent, the concepts elaborated in this
section might also be applied to research in education and welfare and
in some sorts of social research in which there is similarly a hoped for
outcome of direct benefit to the prospective subject. In general, the
3-23
model for this type of research is that defined as innovative therapy
(innovative practices) (1, at p. 9a).
If it is being proposed that the prospective subject consent to
some therapeutic innovation, there should also be at least a disclosure
of what other therapies are available for his condition; this should
include a statement of those that are accepted and approved as well as
others that might be considered innovative. In order to facilitate
rational decision making there should be a reasonable statement of the
risks and benefits of the alternative therapies. Occasionally, this
will demand a frank statement that the risks and benefits of the pro-
posed innovation are relatively less known than are those of the accepted
or approved alternatives. Failure to disclose the existence of alternative
therapies has been held to constitute malpractice (7, at p. 231).
In most cases, if the therapeutic innovation fails it will be pos-
sible to revert to standard and accepted therapy for the same condition.
However, some innovative therapies by their very nature preclude al-
ternative therapies, if so, this should be disclosed. This possibility
may present itself in a variety of situations; some examples follow:
In some diseases there is an ever-present possibility of the sudden
occurrence of some permanent complication. Thus, during the course of
experimentation with a new antihypertensive drug, it may be necessary to
withhold known effective therapy. Should the new drug fail to control the
blood pressure, the subject is liable to the sudden onset of a serious and
3-24
irreversible complication such as cerebral hemorrhage.
Some innovative therapies are designed to permanently ablate some
tissue or organ by either surgery or radiation. Should the therapy fail,
the tissue or organ remains ablated usually with corresponding permanent
loss of function. An innovative program to assist children in over-
coming reading disabilities may prove ineffective. As a consequence,
a crucial period in the child's development during which some other pro-
gram might have had a more or less salutary effect, has been lost irrevo-
cably.
8) There should be: "An offer to answer any inquiries concerning
the procedures" (5, Section 46„3 (c)(5)). Mechanisms for insuring that
the prospective subjects will have a maximum opportunity to have questions
answered are discussed in several subsequent sections.
9) When appropriate there should be a suggestion to the prospective
subject that he might wish to discuss the proposed research with another
before consenting. In some situations, consent of another individual will
be required by regulation; eg, non- therapeutic research directed at the
fetus will require the informed consent of the father as well as the mother
(15). In some other situations consultation with a third party (other than
the investigator and prospective subject) will be either required or strongly
suggested by the IRB. These situations will be discussed in detail below
(cf, Consent Committee and Third Party Scrutiny). In some situations in
which the proposed research entails a consequential amount of risk, dis-
comfort, or inconvenience, to the prospective subject—or, in innovative
therapy, when there are difficult choices between reasonable alternatives- -
3-25
it should be suggested that he might wish either to consult with a
trusted advisor or to share the burdens of decision-making with another.
Most commonly, the trusted advisor will be a "physician- friend".
Ideally, this would be the prospective subject's personal physician
who has no involvement in the research; even more ideally, this physician
would have no affiliation with the institution in which the research is
to be done. When the prospective subject has no personal physician or
when the personal physician is involved in the conduct of the research,
it might be appropriate to offer the services of another physician. In
other cases, depending upon the nature of the research, it might be
suggested that the prospective subject might wish to consult a trusted
clergyman, lawyer, some other appropriate professional advisor, or a
friend (who need not also be a professional).
In some situations the prospective subject might be advised to
involve a third party in the decision-making process even though this
is not required by law or by the IRB. Thus, for example, in studies on
the efficacy of a new post-coital contraceptive: It should be made clear
that since this is an experimental maneuver there is a real possibility
that it might fail. The prospective subject should be advised that she
might wish to have her coital partner involved in the negotiations for
informed consent. This would be particularly true if her partner might
subsequently be called upon to share the burdens of decision-making and/or
financing on the issues of either abortion or carrying the pregnancy to
term should the experimental drug prove ineffective. Of course, she might
refuse to involve her partner in the decision. In this case, she should
3-26
be made aware of all of the potential burdens for which she has assumed
independent responsibility.
Although the author is aware of no case in which a coital partner
has been found not liable for support of a child born after failure of
an experimental post-coital contraceptive, perhaps an analogy might be
drawn from decisions regarding the outcome of artificial insemination
(7, at p. 245). Thus, in the case of Gursky v. Gursky , it was found
that a husband was not liable for the support of a child conceived by
artificial insemination on grounds that he was not consulted and, there-
fore did not consent prior to the artificial insemination. This decision
was rendered on the occasion of the divorce of the two litigants.
A somewhat--but not precisely—analogous situation is presented
when a man is asked to consent to an experimental form of sterilization.
For a variety of reasons, including those mentioned in the preceding
paragraphs, it might be suggested to him that he might wish to share
decision-making responsibilities with his mate. Some states, by statute,
require spousal consent to sterilizations (7, at p. 30). However, even
where such consent is not required by statute, the man with whom consent
for an experimental approach to sterilization is being negotiated should
be made aware of the potential consequences of not sharing the burden of
decision-making responsibilities.
10) It should be stated that the prospective subject is free to
refuse to participate in the research and further, that he is free to
3-27
withdraw from the research at any time. If the research, once commenced,
precludes withdrawal this should be explicitly stated. Further, the
prospective subject should be advised that such refusal or withdrawal
will in no way adversely prejudice his future interactions with his
physician (or other professional), with the investigator, or with the
institution. Assurance that refusal or withdrawal will not adversely
prejudice future interactions is especially required when any relation-
ship exists between the investigator (or any colleague of the investigator)
and the prospective subject which has any potential for coercion; such re-
lationships include physician-patient; employer-employee; faculty- student;
and so on. This element of information is an elaboration of section
46.3(c)(6) of DHEW regulations (5).
A particular concern here is that individuals who are voluntarily
playing the role of patient, student, or employee, or who are applying
to a professional or to an institution for one of those roles, not be
deprived of their rights to enjoy all of the usual expectations of such
roles as punishment for having refused to play the role of subject either
in addition to or in place of the role they wish to play.
In some situations it may be impossible—or at least very difficult--
to make the promises called for in the preceding paragraph. Thus, for some
diseases the only definitive approach to therapy is innovative. For ex-
ample, if a child with a growth defect wishes to receive human growth hor-
mone (HGH) this can be given only to those who agree to participate in re-
3-28
search designed to test its safety and efficacy. HGH, which is in very
short supply, and which is very expensive, is supplied only to investi-
gators who agree to administer it according to a research protocol pres-
cribed by the supplier. Thus, the offer to the parent of a child who
requires HGH that he might refuse to participate in the research without
adversely prejudicing his relationships to the physician- investigator would
be rather vapid. Standard medical care in this situation would consist of
watching the child fail to grow over the years.
A somewhat analogous situation is presented when the patient has a
malignant tumor (or another inevitably lethal disease) for which all
standard modes of therapy have been tried without success. The only de-
finitive approach to therapy might be a therapeutic innovation. However,
in this situation if the patient (prospective subject) refuses to consent
to the research, he may at least be assured that he will continue to re-
ceive all the supportive, palliative, and symptomatic therapy at the
physician's disposal.
When the therapeutic innovation involves a manipulation of the health
delivery system (1, at p. 20) or the introduction of new types of health
professionals (1, at p. 18) it may very often be the case that that parti-
cular institution can offer no alternative. Thus, for example, methadone
maintenance programs may be obliged to state that as a pre-condition of
participation in the program, the patient (subject) is obliged also to
participate in some forms of research. Similarly, experimental clinics
have been established to assist smokers to discontinue this habit; if the
3-29
smoker is unwilling to participate in the research there may be no
alternative facility or service offered by the institution.
A rather novel example was presented by a proposal to video-tape
the interactions of certain individuals on a psychiatric research
ward. The individuals whom the investigators wished to study comprised
approximately one- third of the total population of the ward. However,
it was obvious to all concerned that the video-tapes would also record
the activities of the other two-thirds of the research ward population
(as well as the staff, visitors, and so on). Thus, in conducting negoti-
ations for informed consent with the two-thirds who were not the intended
subjects of the video-tape research, it was necessary to inform them that
their continued participation in their own research projects would also
involve video-taping of their activities. Those who objected to this
did, in fact, have their relationships to the institution prejudiced.
The only way they could escape was to discontinue their participation
as research subjects in studies in which they wanted to be involved.
11) In some studies it is necessary to inform the prospective sub-
ject that some information is being withheld deliberately. This is
particularly true in some types of behavioral and social research in
which disclosure of the purpose of the studies would vitiate the validity
of the results. Ordinarily in such studies it is required that the pros-
pective subjects be informed of all aspects of the study other than its
purpose. There should also be an offer to disclose the purpose at the
conclusion of the study.
3-30
"Undisclosed purpose" research is not confined to the social
sciences. Thus, for example, in biomedical research "patient com-
pliance" studies are dependent for their success upon the agreement
of the prospective subject to remain ignorant of the purpose until
the conclusion of the study. In such cases, the IRB should demand
assurance from the investigators that the information developed will
not be used to embarrass, harass, or otherwise abuse any individuals
who agree to participate. Thus, for example, it would be quite in-
appropriate at the end of such a study to confront a subject with such
statements as: "Now we have caught you; you do not take your medicines"
(patient compliance study). Or: "We have identified you as a sadist"
(social research example). Rather, in general, during the "debriefing"
the subjects should be informed that the study identified in general
the prevalence of non-compliance or tendencies to cruelty.
In the view of the author, deliberate deception-- lying to pros-
pective subjects—should be avoided. The American Psychological As-
sociation indicates that in extraordinary circumstances, deliberate lying
to prospective subjects might be justified (16, at p. 38).
Comprehension
Valid informed consent cannot exist unless the consentor compre-
hends the information upon which his consent is based. In the context
of professional practice our legal system seems to be evolving the con-
cept that the professional not only has the responsibility for informing
a client as to what will be done but also has the responsibility for
3-31
seeing to it that the client understands the information (4, at p. 521).
In the context of medical practice, informed consent has been found not
to exist because patients have not understood such words as mastectomy
(4, at p. 651), laminectomy (7, at p. 232), and so on. Similarly,
patients have been judged to be uninformed owing to "limited command
of the English language" (7, at p. 125). Also in Reyes v. Wyeth Labora -
tories (17) the fact that the plaintiff had "...a seventh grade edu-
cation, but her primary language is Spanish". . .was taken to imply that
she may have "...lacked the linguistic ability to understand..." the
significance of a form she had signed. Parenthetically, in the latter
case, this was not a pivotal point since the form lacked the information
that the court considered important. Such failures in communication can
ordinarily be obviated by explaining technical procedures in lay termin-
ology and by the use of interpreters and translators as necessary.
Incidentally, it should be noted that in some cases the patient
or prospective subject might have a perfectly adequate command of the
English language; the physician, however, may not. This is particularly
a problem when foreign medical graduates are employed as interns, re-
sidents, or post-doctoral fellows. A recent article (18) cites several
trial court cases in which hospitals were found liable for malpractice
because—owing to inadequate command of the English language--a foreign
medical graduate had either taken an inadequate history or failed to pro-
vide instructions to patients.
Additional steps may be taken by the investigator to assure himself
3-32
that the prospective subject comprehends the information. He might,
for example, ask the prospective subject some questions about the
proposed research. He might ask the prospective subject to explain
some of the more important points of information in the subject's own
words. It has been proposed by Miller and Willner (19) that in some
research projects—particularly those in which there is a large amount
of complex information presented to the prospective subject — that this
process might be formalized in what they call the "two-part consent
form 1 '. The first part of the two-part consent form is a standard
consent form prepared for the research project. After the information
has been presented to the prospective subject and after he has had the
opportunity to have his questions answered, he is presented with the
second part of the consent form. The second part consists of a brief
quiz on the essential elements of information that have already been
presented. The prospective subject is asked to respond in his own words
to such questions as: How much time will you be spending in the hospital
if you agree to participate in this research? How much time would you
spend in the hospital if you do not participate in this research? Thus,
through the second part of the consent form it might be documented that
the prospective subject has a clear grasp not only of what he is con-
senting to but also of the consequences of his consent.
It has also been proposed that the consent committee might also
contribute to the assurance that the prospective subject comprehends
(cf, Consent Committees).
3-33
Autonomy
DHEW regulations (5) prescribe that a fundamental feature de-
fining Informed consent Is that consentor be "...so situated as to
be able to exercise free power of choice with undue inducement or any
element of force, fraud, deceit, duress, or other form of constraint
or coercion." Implicit in this clause is the requirement or expectation
that the individual who is negotiating informed consent will somehow
determine adequately that the prospective subject is suitably free to
make the choices presented to him. In general, it will be relatively
easy to avoid intentional fraud or deceit; similarly, it will ordinarily
be easy to detect retrospectively situations in which fraud and deceit
were used to secure informed consent. Obviously, in such cases there
is no Informed consent; rather, there is misinformed consent.
The other components of this requirement may be more difficult to
achieve even by the most sincere negotiators for informed consent.
Similarly, it may be most difficult to determine retrospectively whether
a given subject has been inappropriately coerced or constrained.
It is possible to question whether any individual who perceives him-
self as a member of society is totally free to make choices regarding the
disposition of his own body. In one way or another those who exercise
their choices beyond the bounds of the expectations of their groups
might be perceived as antisocial; they may be described by those who
observe their behavior with adjectives ranging from eccentric to socio-
pathic.
3-34
Perhaps the most important category of prospective subject that must
be considered is that of the patient or "sick person". To illustrate the
barriers presented to the achievement of autonomy by having assumed the
role of "sick person", let us examine Talcott Parsons' view of this role
in relation to the physician (4, at p. 203). Parsons discusses the role
of patient and physician in the context of his definition of health and
the overall role of medicine in society. Health is defined in terms of
a given individual's capacity to perform effectively the roles and tasks
for which he has been socialized, and the concept of a person's health is
integrally defined with respect to a person's participation in the social
system.
Illness is also defined within the context of the social system and
therefore, is judged as being indicative of a disturbance of the capacity
to perform roles and tasks effectively. Parsons identifies four aspects
of the institutionalized expectation system relative to the sick role:
1) There is an "...exemption from normal social role responsibilities,
which of course is relative to the nature and severity of the illness. This
exemption requires legitimation by and to the various alters involved and the
physician often serves as a court of appeal as well as a direct legitimati-
zing agent. . .being sick enough to avoid obligations can not only be a right
of the sick person but an obligation upon him...."
2) "...the sick person cannot be expected by "pulling himself together"
to get well by an act of decision or will. In this sense also he is exempted
from responsibility—he is in a condition that must "be taken care of." "...
the process of recovery may be spontaneous but while the illness lasts he can't
"help it". This element in the definition of the state of illness is crucial
as a bridge to the acceptance of "help"."
3) "...the state of being ill is itself undesirable with its obligation
to want to "get well." The first two elements of legitimation of the sick
role thus are conditional in a highly important sense. It is a relative legi-
timation so long as he is in this unfortunate state which both he and alter
hope he can get out of as expeditiously as possible."
3-35
4) There is an obligation upon the sick person "...to seek
technically competent help, namely, in the most usual sense, that
of a physician and to cooperate with him in the process of trying
to get well. It is here, of course, that the role of the sick person
as patient becomes articulated with that of the physician in a comple-
mentary role structure." (emphasis in the original).
Parsons describes in considerable detail the practical consequences
of this definition of the sick role (20). The second component of the
sick role may be perceived by the sick person as a type of personal
gain. At least temporarily, he is relieved of various duties and ob-
ligations which he might consider onerous. However, this social de-
finition of illness imposes upon the "responsible" sick person the ob-
ligation to seek and cooperate with competent help (ordinarily health
professionals); if he fails in these responsibilities, his sick role
will eventually come to be seen as illegitimate (irresponsible).
A second barrier to the achievement of autonomy is the assumption
on the part of both the sick person and the health professional that the
former--unless he is himself trained in the profession—has no possibility
of achieving the degree of understanding that permits the latter to make
health-related decisions relatively easily. Thus many studies have in-
dicated that sick persons are likely to "go along with" the judgment of
their physician without probing the bases for such judgments. On the other
hand, many physicians have expressed exasperation and frustration with the
prospect of being obliged to Inform a patient to the extent that the latter
can make a rational choice. An illustrative example of the intensity of
such feelings is presented by P.J. Burnham as a model form for consent to
3-36
herniorrhaphy (4, at p. 658).
In reflecting on these and other limitations of the sick role Jonas
(21) reached the tentative conclusion that as we consider who is eligible
to be recruited ("conscripted") for the role of research subject: "Least
and last of all the sick- -the most available of all as they are under
treatment and observation anyway." Yet, he recognizes that "...the very
destination of medical research, the conquest of disease, requires at the
crucial stage trial and verification on precisely the sufferers from the
disease, and their total exemption would defeat the purpose itself." Thus,
subsequently, he proposes: "...the emphatic rule that patients should be
experimented upon, if at all, only with reference to their disease . Never
should there be added to the gratuitousness of the experiment as such the
gratuitousness of service to an unrelated cause."
Katz (4, at p. 727) has observed: "While a volunteering subject can
be alert to protect his own self-interest, a patient-subject's need for
treatment may cause him to overrate the benefits and underestimate the
risks of a research technique." This statement was made in the context of
having defined the volunteering subject as one who has not assumed the "sick
role" while the patient-subject is one who has. May we assume that the vol-
unteering subject is more autonomous than the patient-subject? Perhaps not.
Lasagna and von Felsinger (4, at p. 623) conducted psychological tests on 56
"healthy young male volunteers", all of whom had participated in one or more
clinical drug trials. These young men had been drawn from a college student
population. In this group they found that 25 individuals had manifestations
3-37
of serious psychological maladjustment including three psychotlcs, twelve
psychoneurotics, and so on. Iliey further attempted to examine what factors
motivated these "normal young men" to participate in the research. Many
of the motivations might be viewed by detached observers as "illegitimate":
eg, a search for "thrills" or "kicks"; escape or release from personal pro-
blems and drives; temporary relief from the boredom or pressure of everyday
life; satisfaction of self-destructive urges; and so on.
Some sorts of individuals that might be assumed to have their autonomy
limited have been identified earlier (2, at p. 51). In general, these are
individuals who have been designated as having limited capacities to consent,
having subordinate relationships to the investigator or his institution, or
having--by virtue of other aspects of their life situations--especial vul-
nerability. One particular class of especially vulnerable individuals whose
autonomy it will be most difficult to decide how to preserve is the dying
person. How we might deal with especially vulnerable persons remains to be
seen. To some extent, this depends upon whether the role of subject will be
viewed as a right, a job, or a duty (2, at p. 56).
For the time being we must acknowledge the impossibility of determining
with absolute assurance the prospective subject's freedom to make choices.
The sensitive investigator will--to the extent possible—attempt to de-
termine limits to autonomy through interactions with the prospective subject.
In the section on informing, the issues discussed as elements of in-
formation numbered 1, 3, and particularly 9 and 10 are designed to assist the
individual who is negotiating with the prospective subject to make decisions
3-38
regarding the subject's autonomy (cf, Third Party Scrutiny).
Consent
Ordinarily, after the processes of informing, assurance of compre-
hension, and assurance of autonomy have been completed to a mutually
satisfactory extent, the prospective subject signifies his willingness
to become a subject by consenting. This process most commonly involves
the signing of a consent form; however, in some cases written documentation
of consent may be found unnecessary or undesirable (infra). Thus, it may
be said that at this point a sort of contract has been established. Yet,
as noted earlier, this type of contract is different in several important
respects from the common commercial contract. Some of the differences
have been mentioned earlier. Now another important difference emerges.
The process of consent can not be viewed as merely the consummation
of a contract. The subject has already been informed (element of in-
formation number 10) that he is free to withdraw from the research at any
time and that such withdrawal will in no way adversely prejudice his future
interactions. . .and so on. In the context of medical practice it is clearly
established that a patient may revoke or withdraw consent before or (if
possible) during the medical procedure to which he has consented (7, at p.
261). This principle should certainly be at least equally applied to re-
search. Thus, even after consent has been given, it is necessary to remain
continually aware of the subject's continuing willingness to continue. In
research activities requiring a high degree of commitment of the investigator's
time or resources to a particular subject, as the research progresses, there
is a progressively increasing motivation to the researcher to persuade a
3-39
subject to continue. The Investigator must be aware of his own moti-
vations and take care not to subjugate the will of the subject of his
own (cf, Consent Committees and Third Part Scrutiny).
Standards for informed consent to research and practice
There seems to be nearly universal agreement that negotiations for
informed consent to the investigator- subject relationship should meet
higher standards than those for the physician-patient (or any analogous
professional-client) relationship. This expectation for higher standards
is reflected in most ethical codes and in the views of most commentators
on the subject. However, there have been some recent departures from this
viewpoint. Thus, for example, Capron (4, at p. 574) has asserted: "...the
standard approach has it backwards. Higher requirements for informed con-
sent should be imposed in therapy than in investigation, particularly when
an element of honest experimentation is joined with therapy." Capron
further points out that "...patients who are offered new therapy often have
eyes only for its novelty and not for its risks."
In a subsequent publication (3) in the specific context of discussing
consent to Interventions for catastrophic diseases, Katz and Capron suggest
that it makes little difference whether the intervention is formally clas-
sified as research or practice; the same sorts of information and other as-
pects of the negotiation will be determined by the characteristics of the
proposed intervention.
Feinstein (22) has observed that there is indeed a "double standard"
for informed consent in routine clinical practice and in clinical drug trials
3-40
(innovative therapy). He points out that: "An act that receives no
special concern when performed as part of clinical practice may become
a major ethical or legal issue if done as part of a formally designed
investigation." He expresses the view that demands for formality in
the consent negotiations should be determined by the "architecture of
the clinical research". Thus, he distinguishes two types: 1) "Explicatory"
the motive of which is to explain the mechanisms by which nature works, and
2) "Interventional" the motive of which is to change what nature has done
or to thwart what it might do.
He argues that: "The demand for atotally informed consent seems
reasonable whenever someone becomes a subject in an explicatory experiment.
Since the imposed maneuver is personally unsolicited, medically unnecessary,
and clinically non-beneficial, the potential subject should receive a com-
plete disclosure of what is planned and what might happen."
However: "For each patient who enters a controlled clinical trial...
the imposition of an experimental maneuver is personally solicited, medi-
cally necessary, and intended to provide clinical benefit. The investi-
gator is acting as a doctor practicing medicine; and he engages in a
doctor-patient relationship with each "subject". The only difference
between the trial and other acts of clinical therapy is that the investi-
gator, uncertain about which treatment is best to use, has decided to make
the therapeutic decision for each patient in a pre-planned (usually randomi-
zed) rather than ad hoc judgmental manner."
Thus, he proposes, that the negotiations for informed consent in in-
terventional clinical research (innovative therapy) may be allowed to con-
form more to the norms of medical practice than to those of explicatory re-
search. Parenthetically, there are presented here only the major points
providing the rationale for Feinstein's position; a more fully developed
argument may be found in his article.
In this paper, the author takes the position that patients (or other
clients of professionals) are entitled to the same degree of thoroughness
3-41
of negotiations for informed consent as are subjects. However, for
reasons stated earlier (1, at p. 3) the patient (client) should, in
general, be allowed more freedom than the subject to relinquish this
entitlement. In other words, patients may be offered the opportunity
to delegate decision making authority to a physician while subjects (of
any experiment bearing any consequential possibility of harm) should
rarely be offered this option. The most important distinction between
the negotiations for informed consent in the two contexts (research and
innovative therapy as opposed to practice) is that the prospective subject
must be informed that if the proposed activity is research or innovative
practice, the subject will be at least in part a means and perhaps only a
means to another end (1, at p. 3).
In general the negotiations for informed consent to innovative practice
will be more complex than those to pure research where there is no intent
to bring direct health related benefit to the subject. The greatest pro-
blems will be making clear to the prospective subject the extent to which
he, as opposed to others, might be expected to reap the benefits of the
proposed activity.
It is commonly asserted that, while there are many cases in which
physicians have been found negligent for having failed to provide full dis-
closure in the context of medical practice, there is but one case in which
an investigator was found negligent on the same grounds: Halushka v. Uni -
versity of Saskatchewan (4, at p. 569). Yet, examination of several medical
3-42
malpractice cases found against physicians for failure to provide full
disclosure reveals that what was not fully disclosed is that the pro-
cedure used was experimental, novel, or innovative; eg, Slater v. Baker
and Stapleton (4, at p. 526), Natanson v. Kline (4, at p. 529), Florentine
v. Wenger (4, at p. 529).
Supervision of the negotiations
Consent committees
It has been suggested by several individuals Lhat--at least in some
sorts of research — each of the four component processes of informed con-
sent might be more effectively accomplished if the entire process were to
be observed --and, perhaps, participated in--by some third party. This
third party might be either an individual or a group charged with the
responsibility to see that the best interests of prospective subjects
were served. The first formal proposal to establish such a group may be
found in the DHEW draft proposal (13) for protection of those human subjects
defined as having "limited capacities to consent"; eg, children, the abortus,
the fetus in utero , products of in vitro fertilization, prisoners, and the
institutionalized mentally infirm. In the introduction to this draft it is
stated that :
"Protection Committees are to be established. . .to provide "supplementary
judgment" concerning the reasonableness and validity of the consent given by,
or on behalf of, subjects. The intent of this policy is that institutions
which apply for DHEW funds or submit research in fulfillment of DHEW re-
gulations, must be in compliance with these special protections, whether or
not particular research, development, or demonstration activities are Fed-
erally activities (sic)."
3-43
A definition of "supplementary Judgment" is provided in the same
draft. A detailed description of the composition and duties of the
protection committee for each category of research is further provided.
In a subsequent draft (14) of proposed policy DHEW "...proposes
an extensive revision in this innovative concept. Initially, it ack-
nowledges that the term "protection committee" is pejorative and pro-
poses the term "consent committee" as more appropriate and consistent
with the primary purpose of such bodies. Further, it proposes to
eliminate specific requirements for the size and composition of such
committees".
The introduction proceeds to point out that much of what had been
specified earlier will be made a matter of local (applicant institution)
option subject to approval at the departmental level by the Ethical Ad-
visory Board. Consent committees continue to be proposed for all cate-
gories of research described in this proposed policy (all categories men-
tioned for the earlier draft with the exception that the revised proposal
for children was "reserved" for subsequent publication).
The proposed duties of the consent committee include the following:
"...to oversee the actual process by which individual subjects are selected
and consents. . .are secured, to monitor the progress of the activity (in-
cluding visits to the activity site on a regular basis) and the continued
willingness of the subjects to participate, to intervene on behalf of one
or more subjects if conditions warrant, and to carry out such other duties
as the Secretary may prescribe."
Further, other duties of the consent committee may include:
"(1) Participation in the actual process by which individual subjects
are selected and their consents secured to assure that all elements of legally
effective informed consent. . .are satisfied. . .this might require approval by
the committee of each individual's participation as a subject in the activity
or it might simply call for verification (e.g., through sampling) that pro-
cedures prescribed in the approved application or offer are being followed."
3-44
"(2) Monitoring the progress of the activity and the continued will-
ingness of subjects to participate." This might include: "...identifi-
cation of one or more committee members who would be available for con-
sultation with subjects at the subjects' request, continuing evaluation
to determine if any unanticipated risks have arisen and that any such risks
are communicated to the subjects, periodic contact with the subjects to
ascertain whether they remain willing to continue. .. .providing for the
withdrawal of any subjects who wish to do so, and the authority to terminate
participation of one or more subjects with or without their consent where
conditions warrant."
(b) The size and composition of the consent committee must take into
account such factors as:
"(3) Whether the membership has been so selected as to be competent
to deal with the medical, legal, social, and ethical issues involved in the
activity; (4) whether the committee includes sufficient members who are un-
affiliated with the applicant or offeror apart from membership on the com-
mittee; and (5) whether the committee includes sufficient members who are
not engaged in research, development, or related activities involving human
subjects. The committee shall establish rules of procedure for carrying out
its functions and shall conduct its business at convened meetings, with one
of its members designated as chairperson."
There is a provision under which an applicant or offeror may request
of the Secretary a modification or waiver of the requirement for a consent
committee where a particular activity involves negligible risk to the sub-
jects.
It is quite clear that the intent of the proposal to develop consent
committees was to enhance the likelihood of success of each of the four
components of negotiating informed consent. While there is little cause
to question that the existence of consent committees would contribute to
this goal, there is no cause to suspect that their existence would result
in perfection of the negotiations for informed consent. In a position
paper submitted by the Association of American Medical Colleges in testi-
mony to the Commission (24) it is argued that the costs (time, money, and,
3-45
perhaps most Importantly, social costs) of developing consent committees
as described in DHEW proposals might far outweigh the expected benefits
to the process of negotiating informed consent. It is further proposed
that similar improvements might be achieved through far less costly means.
In particular, it is proposed that--at the discretion of the investigator
(whose discretion will be reviewed and perhaps modified by the IRB)--"third
party scrutiny" of the consent process might be offered as an option to
some subjects; in some other cases a requirement for third party scrutiny
might be imposed whether or not the subject (or, for that matter, the in-
vestigator) wishes it.
While the final DHEW regulations governing research on the fetus (15)
make no mention of consent committees, the possibility that they might be
required for other categories of prospective subjects having limited capa-
cities to consent still exists. Therefore, the concept of the consent
committee merits serious consideration by the Commission.
In the opinion of the author, the entire research process and all of
its participants stand to lose more than they might expect to gain from the
establishment of formal consent committees as described in DHEW proposed
policies. The reasons for this position are essentially those detailed in
the above cited position statement (24). In its place, one might consider
extending the use of third party scrutinizers.
Third party scrutiny
The necessity for offering the third party scrutinizer, either as an
option or as a requirement, should be the responsibility of the IRB. The
3-46
third party scrutinizer may be assigned any of the responsibilities
specified for the consent committee in the DHEW proposal.
The types of individuals that might be employed as third party
scrutinizers were identified earlier (Informing, element no. 9). This
individual should be someone whom the prospective subject will recognize
as a trusted advisor. Most commonly, this will be a "physician-friend";
in other cases depending upon the nature of the decision-making problem
and the wishes to the subject it might be a clergyman, a lawyer, a re-
lative, or some other appropriate professional or nonprofessional advisor.
The types of research for which third party scrutiny might be
found useful are those which have some or all of the following features.
The subject may be called upon to assume substantial risk or inconvenience
particularly when there is no expectation of direct health related benefit.
In a proposal to conduct a therapeutic innovation, there may be legitimate
competing alternatives; this presents a particular problem when the choice
of one might foreclose any possibility of reverting to the alternative if
the one chosen should fail. (Informing, element no. 7). There may be some
cause to suspect substantial barriers to comprehension or autonomy (cf,
Comprehension and Autonomy).
In such cases during the negotiations for informed oonsent it may be
proposed to the prospective subject that he might wish to enlist the services
of a third party scrutinizer. He might further be alerted as to the general
types of functions there might be for such a person in this particular re-
search proposal. He might further be advised as to what sorts of persons
3-47
he might wish to choose to assume this role. After receiving all of this
information, ordinarily the prospective subject should have the option to
refuse the services of a third party scrutinizer. Imposition on the subject
of a requirement for a third party scrutinizer is a serious step. It is
a potential invasion of privacy (particularly when the scrutinizer is to
be chosen by the institution, and even more particularly, when the individual
is not an agent or employee of the institution conducting the research (24).
Also, imposing such a requirement will, at times, be tantamount to a declara-
tion to the prospective subject that his judgment (ability to comprehend;
ability or freedom to make choices; and so on) is to be questioned. However,
in some cases this will be necessary.
Illustrative example: There follows an example of how and why an IRB
imposed a requirement for third party scrutiny of a portion of the process
of negotiating informed consent in a particular research protocol. This
example is selected for several reasons. The proposed subject population
was at least potentially vulnerable by virtue of two attributes not identified
in existing or proposed DHEW regulations or policies. Further, the analysis
of risks and of hoped for benefits were both complex.
The research maneuver involved inserting a catheter into the coronary
sinuses for purposes of sampling blood to assay a substance made in the
heart. Under no circumstances was the catheter to remain in the coronary
sinus longer than 10 minutes. The subject population was individuals who
had been admitted to a coronary care unit very recently with known or sus-
pected myocardial infarctions. No subject would be enlisted who did not
require catheterization of the right side of the heart for "practice"
purposes. No "extraordinary" measures would be taken to place the catheter
3-48
in the coronary sinus in individuals in whom such placement was found to
be difficult.
In other words, all possible measures were taken to minimize the
probability of harm. Yet there remained a very small probability of
harm in the form of arrhythmia. The facilities available for monitoring
for the occurrence of this harm were optimal; the probability of success of
interventions proposed to counteract a arrhythmias had a very high likeli-
hood of success.
No direct benefits to the subjects were anticipated. The subjects
would receive the benefit of the catheterization indicated for practice
purposes. As discussed earlier (2, at p. 35) because the period of very
close monitoring would be extended, there was an enhanced likelihood of
detecting various complications (of the myocardial infarction) promptly
and, when possible, initiating definitive treatment. There was a moder-
ately large probability of a very large benefit to the population of in-
dividuals having myocardial infarctions. The investigators were at-
tempting to identify a chemical which might contribute to the morbidity
and mortality of this disease; if they were able to establish the re-
lationship it would be possible to intervene directly; means for blocking
the formation as well as the effects of this chemical already exist. Prior
studies done in animals and in humans not having myocardial infarctions in-
dicated a reasonably high probability of associating this chemical with the
pathogenesis of the disease.
In medical practice it is customary to assume that any stress may be
harmful to individuals with myocardial infarctions. Thus, great care is
taken to avoid physical and psychological stresses particularly during the
3-49
early stages of treatment. Parenthetically, it is not proved that
psychological stress increases the morbidity or mortality in this
disease. Patients with known or suspected myocardial infarctions
are commonly treated with narcotics to diminish pain and anxiety.
Thus, most individuals whom the investigators might approach to negoti-
ate informed consent for this research will have recently received
treatment with a narcotic.
In the judgment of the IRB, the negotiations for informed consent
would create anxiety. This, in turn, might jeopardize the physical and
psychological well-being of the prospective subject. (As noted earlier,
this could not be proved.) Further, it was judged that the proposed re-
search population was especially vulnerable for either of two reasons:
1) Some of these individuals might perceive themselves as in the process
of dying; in some cases this perception might be in accord with facts.
2) Some of these people might have had their abilities to make rational
judgments impaired by virtue of having received a narcotic; ie, a state
of inebriation would have been induced.
Thus, the IRB imposed a requirement for partial scrutiny of the con-
sent process by two different types of third parties:
1) Before any approach to the prospective subject was made, the in-
vestigator would review the proposal- -all elements of information- -with the
next of kin. The purpose of this discussion was to determine if--in the
view of the next of kin--this is the sort of thing to which the prospective
subject might be expected to consent. If in the judgment of the next of kin,
the answer was no, no approach to the prospective subject would be made.
3-50
2) A physician not connected with the research and who, by virtue
of his relationship to the prospective subject, had the best interests
of the prospective subject in mind would be called upon to determine that:
a) The patient's physical and psychological condition were such that
he was not likely to be unduly threatened or harmed by the consent negoti-
ations, and
b) the patient's cognitive function had not been impaired to the ex-
tent that he could not understand the information.
The physician assigned responsibility to make these determinations
was to be the one who had most recently examined the patient sufficiently
thoroughly to make these determinations. When possible this would be the
patient's personal physician who had established a physician-patient re-
lationship prior to the onset of the current illness. When no such re-
lationship existed or when the physician meeting this description was un-
available, the physician called upon for this judgment would be the resident
physician in the coronary care unit. In the event the personal physician
was also a member of the research team, he would be disqualified as a third
party scrutinizer.
In this case, the investigators were cardiologists who are highly
skilled and experienced not only in catheterizations of the heart but
also in the management of patients with myocardial infarction in the
coronary care unit. Thus, it was judged unnecessary to have any other
physicians in attendance during the process of negotiating for informed
3-51
consent or during the brief period during which the research maneuver
was performed. In the event the subject wanted to see his own physician
during this period his access to that physician would be no more limited
than it would ordinarily be in the usual conduct of activities on a
coronary care unit. The circumstances of the research were such that
it would not be in the scientific interests of the investigators to
continue the research if something "went wrong".
Other examples of research in which a requirement for third party
scrutiny of all or part of the negotiations for informed consent can be
provided should the Commission wish them. It is the view of the author
that regulations for third party scrutiny of negotiations for informed
consent should impose on the IRB the requirement for determining when it
will be necessary and what form it should take. Regulations that attempt
to prescribe the form and substance of third party scrutiny may deprive
the IRB of the flexibility necessary to meet the needs of particular re-
search proposals and particular prospective subjects.
Documentation of informed consent
DHEW regulations prescribe in section 46.10 the forms in which in-
formed consent must be documented. Three possible forms are described
(5):
"(a) Provision of a written consent document embodying all of the
basic elements of informed consent. This may be read to the subject or
to his legally authorized representative, but in any event he or his
legally authorized representative must be given an adequate opportunity
to read it. This document is to be signed by the subject or his legally
authorized representative. Sample copies of the consent form as approved
by the Board (IRB) are to be retained in its records.
3-52
"(b) Provision of a "short form" written consent document in-
dicating that the basic elements of informed consent have been presented
orally to the subject or his legally authorized representative. Written
summaries of what is to be said to the patient are to be approved by the
Board. The short form is to be signed by the subject or his legally
authorized representative and by an auditor witness to the oral presen-
tation and to the subject's signature. A copy of the approved summary,
annotated to show any additions, is to be signed by the persons offic-
ially obtaining the consent and by the auditor witness. Sample copies
of the consent form and of the summaries as approved by the Board are to
be retained in its records."
The third "form" is, in fact, not a form; it reads in part as follows
"(c) Modification of either of the primary procedures outlined in
paragraphs (a) and (b) of this section. Granting of permission to use
modified procedures imposes additional responsibility upon the Board and
the institution to establish: (1) That the risk to any subject is mini-
mal, (2) that the use of either of the primary procedures for obtaining
informed consent would surely invalidate objectives of considerable im-
mediate importance, and (3) that any reasonable alternative means for
attaining these objectives would be less advantageous to the subjects."
The regulations go on to require that the reasons for permitting the
use of modified procedures must be individually and specifically docu-
mented and made a matter of permanent record.
Careful reading of the quoted portion of paragraph (c) will indicate
that--since all three criteria must be met—modified procedures may
virtually never be used. If the word "and" preceding "(3)" were changed
to "or", some flexibility might be permitted by this paragraph.
It is difficult to understand the implications of these regulations
without an awareness of how they are implemented by DHEW officials having
such authority. Thus, an illustrative example is provided:
Since these rules and regulations were promulgated it has been NIH
policy to require informed consent documented as specified in either
paragraph (a) or (b) to use tissues or fluids removed at either surgery
or autopsy for research purposes (25). This requirement is imposed even
3-53
though the autopsy or surgical procedures are done in accord with
customary medical practice, where there is no removal or retention
of additional tissues or fluids for research purposes (beyond the amount
required by the medical needs of the patient or next of kin), and where
the procedures themselves are authorized by consent (of the sort usually
used in the medical practice context) to remove, retain, and dispose of
tissues. Further, this requirement is imposed even if there is no way
to link the information that might be developed through the research to
the name of the individual from whom the tissue or fluid was removed. Thus,
it is asserted that current DHEW regulations- -although they seem to permit
reasonable f lexibility--actually permit very little. The consequences of
this sort of interpretation of the regulations will be analyzed subsequently
(Conditions under which consent negotiations may be less elaborate).
The regulations of the Food and Drug Administration do permit some
flexibility (26).
Section 310.102 (a) specifies that "...the use of investigational new
drugs on humans shall impose the condition that investigators "obtain the
consent of such human beings or their representatives, except where they
deem it not feasible or, in their professional judgment, contrary to the
best interests of such human beings."
(b) This means that the consent of such humans (or the consent of
their representatives) to whom investigational drugs are administered
primarily for the accumulation of scientific knowledge, for such purposes
as studying drug behavior, body processes, or the course of a disease, must
be obtained in all cases, and, in all but exceptional cases, the consent of
patients under treatment with investigational drugs or the consent of their
representatives must be obtained."
(d) Defines exceptional cases as "...those relatively rare cases in
which it is not feasible to obtain the patient's consent or the consent of
his representative, or in which as a matter of professional judgment ex-
ercised in the best interest of a particular patient under the investigator's
care, it would be contrary to that patient's welfare to obtain his consent."
3-54
(g) "Contrary to the best interests of such human beings" applies
when the communication of information to obtain consent would seriously
affect the patient's well-being and the physician has exercised a pro-
fessional judgment that under the particular circumstances of this patient's
case, the patient's best interests would suffer if consent were sought."
This section goes on to specify that consent to receive an investi-
gational new drug in Phase I and Phase II shall be in writing. When con-
sent is necessary under such rules in Phase III investigations, it is the
responsibility of investigators, taking into consideration the physical and
mental state of the patient, to decide when it is necessary or preferable
to obtain consent in other than written form. When such written consent
is not obtained, the investigator must obtain oral consent and record that
fact in the medical record of the person receiving the drug.
These quotes from FDA regulations indicate that this agency is willing
to permit more flexibility than is DHEW not only in the documentation of
but also in the nature of the negotiations for informed consent. These re-
gulations might be criticized on grounds that they leave too much to the
judgment of the physician (investigator). However, section 312.1 imposes
the necessity for an "institutional review committee" (analogous to the
IRB) which is assigned responsibility for, among other things, imposing
group judgment on the plans for negotiations for and documentation of in-
formed consent.
The provision in FDA regulations that in exceptional cases, it might
be appropriate to conduct phase III drug trials according to reduced
standards of consent recognizes that in some situations according to the
judgment of the physic ian- investigator , negotiating informed consent might
be contrary to the welfare of a particular patient. In general, this pro-
vision is meant to recognize two sorts of contingencies. The first is the
3-55
situation in which a Phase III drug must be administered immediately in
order to have optimal effect. If time is taken to negotiate informed
consent, the patient-subject might sustain irreversible injury as a
consequence of the delay. It is not extraordinary that physic ian-inves-
and receive from
tigators ask of the IRB permission to proceed with sufficient flexibility
to accommodate the demands of an emergency situation.
The second is the situation in which the patient-subject might for
various reasons not wish to know some of the facts enumerated as elements
of information. Most commonly, this calls to mind the patient with some
dread disease who might be harmed psychologically by inappropriately pre-
mature full disclosure of the diagnosis. However, there might be many
other points of material information with which the patient-subject might
not wish to be bothered. As Louisell and Williams have observed in the
context of medical practice (27, at. p. 594.63): "...Some patients do
prefer not to know the details, including the risks, of proposed pro-
cedures. Their attitude frankly is: "Doctor, I engaged you because of
my confidence in your judgment and skill. I cannot understand all the
implications of what you propose to do, and I do not wish to try. Because
of my confidence in you, and my own ignorance and fears, I ask you to bear
on your shoulders not only the medical burden of the procedure, but the
moral one of the decision." If that is the true attitude of the patient,
and he clearly conveys it, there is no reason in law, morals, or public
policy why the physician normally should not abide it, if he is willing
to do so."
Thus, in the context of medical practice, it seems to be the right
of the patient to waive any element of informed consent he wishes. However,
3-56
as discussed elsewhere, the subject ordinarily should be much more
cautious about delegating decision-making responsibility to the in-
vestigator. The author can recall no proposal by an investigator to
the IRB at Yale University School of Medicine to withhold information
from a prospective subject because--in the judgment of his physician--
full disclosure (particularly of the diagnosis) would be contrary to
his best interest (except, as noted earlier, in cases of emergency).
The consent form
In almost every research activity involving consequential risk
or inconvenience to the subject it is most appropriate to document in
writing the fact that appropriate negotiations for informed consent
have been conducted. Exceptions will be specified in subsequent sections.
In the experience of the author, the consent form used by virtually
every investigator is of the sort specified in section 46.10 (a) of the
DHEW rules and regulations. The view prevails among investigators that
the "short form" prescribed in 46.10 (b) offers no advantage to either
investigator or subject. If the same quality and quantity of information
is to be presented to the subject in form (b) as in form (a), and if all
this information is to be committed to paper no matter which form is used,
there seems no advantage to not giving the subject a copy of that paper
to read, sign and--as discussed subsequently- -keep. The author recalls
no case in which an investigator proposed to the Yale Medical School IRB
to use a "short form" as described in the regulations. Through the
Commission's IRB study it will be seen if this experience is shared by
3-57
most or all other IRBs.
In general, it should be kept in mind by all concerned- -but parti-
cularly the investigator and members of the IRB--that the consent form
can almost never be so constructed as to anticipate all of any parti-
cular prospective subject's wishes to be informed. The consent form
is most effective when it is viewed by the investigator as an instru-
ment designed to guide him in his negotiations with the prospective sub-
ject. The consent form should contain at least the minimum amount of
information and advice that should be presented during the negotiations
for informed consent. If any substantive new understandings are developed
in the process of the negotiations which have any bearing on the prospective
subject's willingness to participate these should be added to the consent
form signed by that particular individual.
The consent form should present an adequate coverage of each of the
eleven elements of information (section on Informing) that are germane to
the particular proposed research activity. The fact that the consent form
is considered a guide to the negotiations should be specifically reflected
as indicated in element number 8 calling for an offer to answer any in-
quiries concerning the procedures. It should be made clear that this
offer to respond to inquiries is, in fact, an offer to elaborate on any
of the eleven elements of information to the extent that the prospective
subject wishes.
The use of general consent forms designed to document consent for
research generally or even for several categorically related research
protocols should be discouraged. At the very least, a consent form should
3-58
be designed to meet the specifications of a particular research proposal.
A fully satisfactory document designed to meet the needs of one particular
protocol may, as discussed above, undergo further modification to reflect
alterations in the negotiations that occur as a consequence of discussing
the proposed research with individual subjects. At times it is necessary
to design more than one consent form for use in a single research protocol.
Thus, for example, when a protocol is designed to conduct the same maneuvers
on two distinctly different populations of subjects — eg, diabetics and
healthy volunteers- -it is ordinarily appropriate to have separate consent
forms for each class of subject. Many of the elements of information pre-
sented to the diabetics will differ from those presented to the healthy
volunteers. Thus, for example, element number 3 would contain very different
information as to why the prospective subject has been selected. It might
also contain very different sorts of information regarding "pretests" to
determine eligibility and the consequences of "failure to pass" these exam-
inations. Further, the nature of the benefits to be derived is likely to
be very different.
Many institutions have guidelines prescribing the type of language
that ought to be used in the consent form. Most commonly it is suggested
that the form be worded in language that the "average lay person" should
be expected to understand. This suggestion has provoked considerable
controversy. For example, what is an average lay person? And what can
he be expected to understand? Many protocols are designed to involve
subjects that depart in some substantial way from "average". For example,
some studies are designed to involve subjects who have serious chronic
diseases in whom "standard and accepted" therapeutic measures have failed.
3-59
Commonly, such individuals have a highly sophisticated understanding of
the technical language used to describe their disease, various means
of diagnosis and therapy, and the harms that may occur as a consequence
of therapy or of not being treated. On the other hand, some protocols
are designed to involve naive subjects who might have little schooling,
who might have primary languages other than English, and so on. Thus,
it seems more appropriate to suggest in guidelines that the consent
form should be presented in language that the prospective subject might
be expected to understand. Protocols in which it is planned to draw
upon diverse populations for subjects might require more than one con-
sent form.
Various institutions have developed further recommendations on the
language of the consent form. Some institutions have policies requiring
that the entire consent form be worded in the first person. Thus, the
various elements of information each begin: "I understand that the
purpose of the study is to... "or "I hereby agree to have Dr. Jones draw
10 cc (2 teaspoons) of blood...". Other institutions recommend that
consent forms be worded in the second person. For example: "You are
invited to participate in a research project designed to. .. (accomplish
some purpose)." "If you agree to participate there is a small possibility
that you might develop a rash." "The purpose of this research is not to
bring direct benefit to you but rather to develop information which might
help us design better diagnostic methods for persons like you in the future."
The preference of the author is for consent forms that present the
elements of information in the second person. After the elements of in-
3-60
formation have been presented there may be a statement worded in the
first person entitled " Authorization : I have read the above and de-
cide that... (name of subject) will participate in the project as des-
cribed above. Its general purposes, the particulars of involvement,
and possible hazards and inconveniences have been explained to my
satisfaction. My signature also indicates that I have received a copy
of this consent form."
In the view of the author, the presentation of the information in
the second person followed by an authorization written in the first
person best conveys the sense of negotiation (of give and take) in-
volved. However, the author is not aware of any documentation that
one style of language is to be preferred to another.
At the conclusion of the consent form there should be a line pro-
vided for the signature of the consentor. There should further be some
means of specifying how the consentor is related to the subject (self,
parent, guardian, and so on). It should also specify the date on which
the form was signed. Some individuals may wish to also record the date
on which the form was first presented to the prospective subject if it
differs from that on which it was signed. In cases where proxy consent
is obtained a space may be provided for the signature of the actual subject
if appropriate. This signature may indicate, depending upon the situation,
the actual subject's consent (which may or may not be legally valid), his
assent, or, perhaps, merely his awareness that somebody is consenting to
something on his behalf.
Also on the consent form there should be a space provided for the
signature of the individual who has conducted negotiations for consent.
His telephone number should be provided in the event the prospective subject
3-61
(or subject) has any wish to discuss the project further. In cases where
a third party scrutinizer has participated in the consent process, a space
should be provided for his signature as well. In addition, it might be of
value in some consent negotiations to provide the names and telephone
numbers of other individuals who might be consulted by the prospective
subject or his agents for further information. The most common situation
in which it is advantageous to list the name(s) and telephone number (s)
of other individuals is when consent for a complex research activity is
obtained by someone other than the professional who is responsible for
the conduct of the activity (cf, Who should negotiate for informed consent?)
Prohibition of exculpatory clauses
DHEW regulations (5) require under section 46.9 that: "No, . .informed
consent, oral or written, .. .shall include any exculpatory language through
which the subject is made to waive, or appear to waive, any of his legal
rights, including any release of the institution or its agents from lia-
bility for negligence." The author finds no cause to elaborate, modify,
or challenge this requirement as being applicable to all negotiations for
informed consent.
Who keeps records of informed consent ?
There seems to be general agreement that the IRB should keep copies
of the general consent forms offered to each prospective subject of any
particular activity. A copy of the consent form (or the various consent
forms) used to document the negotiations for informed consent is made
part of the permanent record of each protocol that is retained by the IRB.
There have been various proposals that the IRB should also keep copies of
the consent forms signed by each subject. The author suggests that this
3-62
would in some cases be detrimental to the best interests of the sub-
jects; the reasons for this suggestion are the same as those presented
for avoiding having the signed consent forms made a part of the permanent
records of the institution ( infra ) .
The investigator should ordinarily assume the responsibility to
retain the consent forms signed by the individual subjects. It should
be his responsibility to safeguard the confidentiality of these forms
when appropriate and to the extent necessary. To some extent the mere
fact that an individual has agreed to serve as a subject of a particular
research activity will reveal information about that subject that he
might not wish to have generally known. Thus, for example, the consent
form might specify: "You are invited to participate in this research
project because you have cirrhosis of the liver." Various rules and
regulations specify how long the investigator ought to retain these
consent forms in his files. Since one of the purposes of documentation
of informed consent is to protect the investigator from inappropriate
litigation for failure to provide "full disclosure" it is recommended
that the signed consent form be held for a period of time slightly
greater than the statute of limitations for such actions in his juris-
diction. Ordinarily, signed forms should be retained for a sufficient
period of time longer than the statute of limitations as usually ad-
vised for physicians who are concerned with potential malpractice liti-
gations; this is usually about six months but, in states having "dis-
covery statutes", it may be forever (7, at p. 322).
3-63
Many institutions have policies that signed consent forms should
be made part of the permanent records of the institution. This is
based upon three assumptions: 1) If there is any litigation against
the investigator, the institution is likely to be named as a co-de-
fendant. 2) Institutional administrators commonly believe that the
record keeping systems of the institution are superior- -in various ways--
to those of investigators. 3) Investigators may move away from the in-
stitution to take employment elsewhere. In such cases, the signed con-
sent forms might not be available to the institution when they are
needed. Each of these assumptions is at least in part true. The only
one that might be seriously challenged by some investigators is the
second. Some investigators are more adept at keeping orderly records
than almost any institution. The relatively smaller number of records
they keep as well as the relatively smaller number of individuals who
have access to these records each contribute to their lesser likelihood
of loosing a record.
On the other hand, individual investigators or small groups of in-
vestigators are much more likely to be able to provide assurance of con-
fidentiality than are large institutions. In general, when confiden-
tiality is a significant issue, the author recommends that the investigator
be assigned primary responsibility for keeping signed consent forms.
In most cases the subject should be given a copy of the consent
form bearing all the signatures that are to be placed on that form.
Since, as discussed earlier, this is a form of contract, it seems only
appropriate to give ttfe subject a copy. Also, providing the subject with
3-64
a copy of the form will afford him an opportunity to continue to get
more information as additional questions occur to him. It will also
be available as a constant reminder of his freedom to ask questions,
to withdraw without prejudice, and so on. Finally, as suggested above
(cf, Consent form) the form may provide the name(s) and telephone
number (s) of individuals he might wish to contact as necessary.
Who should negotiate with the prospective subject ?
Some institutions have policies prescribing more or less speci-
fically who may negotiate with the prospective subject for informed
consent. Thus, for example, some hospitals require that if the prin-
cipal investigator on a research activity is a physician, the responsi-
bility for negotiating for informed consent cannot be delegated to a non-
physician. Some students of or commentators on the informed consent
process have suggested that the quality of informed consent might some-
how be reduced by virtue of the fact that a physician- investigator dele-
gated responsibility for informed consent negotiations to a research-nurse
or, perhaps, even to a non-professional (28, at pp. 212 et seq) . The
question thus arises as to whether there is any need for regulations
defining who may participate in the consent negotiations either generally
or in relation to particular classes of research. After weighing the pros
and cons of having various types of individuals negotiating for informed
consent, the author will conclude that a good general guideline might
be developed as an analogy to the laws governing the authority of the
physician to delegate responsibilities to various individuals based upon
3-65
his judgment that they are sufficiently trained or experienced to perform
the responsibility adequately and that the physician is held accountable
for the actions of those to whom he delegates responsibility. In general,
these statutes are written to control medical practice and, further, are
written to permit delegation of authority to various types of "physicians
trained assistants". However, It la now proposed that tills model might
be applied to the negotiations for informed consent in a research context
and, further, that the type of professional might be extended to include
all sorts of qualified investigators in addition to physicians.
In general, as one examines the various component processes of the
negotiations for informed consent, it might be assumed that various types
of professionals might be better equipped to accomplish each of these
purpo8es--either by virtue of their training or expertise or by virtue
of their motivations- -than others. The principal investigator will
ordinarily be better informed about the technical aspects of the research
than will other members of the research team. Thus, if he is willing, he
will be more capable of responding to detailed questions about risks,
benefits, alternatives, and so on than will most others. Professionals
who are equally certified might be expected to have approximately the same
amount of detailed information at their disposal.
On the other hand, in the context of medical practice, particularly
in large institutional settings, it has been observed that patients tend
to be better informed when a professional who is not a physician is as-
signed responsibility for their "education". It seems that such professionals
as clinical pharmacists, physician's assistants, nurse practitioners, and so
3-66
on, are generally more oriented toward patient education than are most
physicians. There are, of course, clear individual exceptions to this
generalization. However, there is no a priori reason to assume that
subjects will in general be better informed merely because it is the
principal investigator who undertakes the responsibility for informing
them.
Much has been written about barriers to comprehension that are
created when the individual negotiating with the prospective subject
is of a very much different social class or has a very much higher degree
of education than the prospective subject. Perhaps, it might be reasonable
to suppose that the more equal the negotiators are in these two categories
the more likely there is to be comprehension. In addition, relative equals
would probably be more capable in perceiving non-verbal manifestations of
non-comprehension.
As discussed earlier (cf, Autonomy), when the prospective subject has
assumed the "sick role" there immediately exists a barrier to his autonomy.
In this paper it is intended to expand the definition of "sick role" beyond
the confines of what is ordinarily termed a patient. The very same sorts of
problems exist when a person assumes analogous roles in educational or social
contexts. In these contexts, because of the social demand upon the "sick
individual" to get well, he may find it difficult to refuse to "cooperate"
with a professional whom he believes has the power to "make him well".
It is concluded, that there is no a priori reason to predict that
individuals other than the principal investigator will be more or less suc-
cessful in negotiating (as contrasted with securing or obtaining) informed
consent. It is suggested that the individual (s) assigned this responsibility
3-67
within a research team be the one(s) who seems most capable and most
interested in performing this role. In the event this is a person
having professional qualifications or certification lower than that
of the principal investigator this should be made clear during the
consent negotiations. In such cases it is ordinarily advisable to
name the principal investigator (or one of his equally certified co-
investigators) on the consent form and to provide instructions as to
how he might be made available to respond to questions that--in the
view of the prospective consentor- -cannot be answered adequately by
the individual with whom he is negotiating. The principal investigator
should be held accountable for the actions of all individuals to whom
he delegates responsibility. The selection of the individual (s) who
will function as negotiators for informed consent in any particular re-
search project should be determined by the principal investigator. The
IRB should review this designation and make recommendations or impose re-
quirements for modification as appropriate.
Access to prospective subjects
In some research proposals a prospective subject population will be
identified as one to which the investigator would not have access in the
course of his customary professional activities. Thus, the investigator
may wish to secure the names of prospective subjects through contact with
practicing professionals (eg, physicians, social workers, and so on) or
through the records of physicians, hospitals, schools, welfare agencies,
3-68
and so on. In many cases the clients of these professionals or in-
stitutions may assume that their relations with the professionals or
institution are, at least in part, confidential. Thus, when this mode
of access to a prospective research subject is to be used, it ordinarily
becomes necessary to involve the practicing professional or his in-
stitutional record system in the consent process in a way that minimizes
the potential for violations of confidentiality or coercion.
One mechanism that is commonly used is to have the investigator in-
itiate contact with the practicing professional. The investigator may
provide the practicing professional with a form letter which might be
addressed to suitable prospective subjects. This letter might describe
in general terms the proposed research activity and suggest to the
prospective subject that should he have any interest in participation,
he might initiate direct contact with the investigator . .At that point
negotiations for informed consent might commence as in any other research
context. When appropriate, the prospective subject should be advised in
the communication from the practicing professional that not only is there
no need to inform the practicing professional whether the client consents
to become a subject, but also (when appropriate) the investigator will
not reveal to the practicing professional whether an investigator-subject
relationship has been established. Thus, to the extent necessary, it is
possible to minimize potentials for violations of confidentiality as well
as potentials for coercion or duress. The latter potential presents it-
self most significantly in research designed to develop new knowledge on
some characteristic of the professional-client interaction.
3-69
In some other situations the prospective subject may be the client
of a professional whose overall "care" of the client is dependent upon
his awareness of all activities that might have any influence on the
"health" of the client. In such circumstances, no matter how access to
the prospective subject is established, negotiations for informed con-
sent should not proceed until the consent or approval of the practicing
professional has been obtained. In such cases, full disclosure would
demand that the prospective subject be informed that the research is
proceeding with the awareness of the practicing professional who will
be kept informed of any consequential findings, adverse reactions, and
so on. In general, in most hospitals, there exist policies requiring
that the personal physician of the patient be asked to approve the in-
volvement of the patient in a research activity.
As discussed above (section on Third Party Scrutiny) the practicing
physician may be called upon to serve the role of third party scrutinizer.
Timing of the negotiations for informed consent
To the extent possible and necessary adequate time should be pro-
vided to accomplish the objectives specified under elements of information
numbered 4e, 8, 9, and in the sections entitled Comprehension, Autonomy,
and Third Party Scrutiny.
In addition, in many types of research proposals it is planned to
do the research on individuals who will be relatively incapacitated,
distracted, or preoccupied at the time the research will be done. Thus,
for example, if it is planned to investigate an innovative approach to
delivering a baby, it is possible to identify a suitable population of
3-70
prospective subjects weeks or months before the research will be
done on any particular subject. Ordinarily, it would be preferred
to negotiate informed consent during or after a routine pre-natal
visit with the obstetrician rather than when the woman is experiencing
the discomfort and anxiety of labor and, perhaps, has been already
treated with drugs that might influence her cognitive function. In
general, the obstetrician's office is an environment more conducive
to rational decision making than is the labor or delivery room (28).
Also, prospective subjects contacted during a routine pre-natal visit
will have the time to accomplish the various functions specified in
the preceding paragraph.
In situations in which informed consent has been negotiated well
in advance of the proposed research, the prospective subject should
be reminded of his right to withdraw without prejudice shortly before
the research maneuver is initiated.
In some other circumstances, it will be impossible to conduct the
consent negotiations at a time and in an environment that is most con-
ducive to high quality decision making; see for example, the illustrative
example provided in the section on Third Party Scrutiny.
Conditions under which consent negotia tions may be less elaborate
Thus far, this paper may have created the impression that the
average negotiation for and documentation of informed consent must be an
extremely elaborate process. This false impression derives from the fact
that an attempt has been made to present a comprehensive account of the
3-71
various factors that must be considered by and the various devices
available to individual (s) who are planning negotiations for informed
consent. In fact, in most cases most of these factors and devices will
be found inappropriate or unnecessary. Each negotiation for informed
consent must be adjusted to meet the requirements of the specific pro-
posed activity and, more particularly, be sufficiently flexible to meet
the needs of the individual prospective subject. By way of illustration,
most consent forms approved by the IRB at Yale University School of
Medicine contain less than one half page of single-spaced text. Sample
consent forms published by the UCSF Committee on Human Research (29) con-
tain approximately the same amount of information.
Ingelfinger in an article entitled "The Unethical in Medical
Ethics" has expressed alarm over the increasing apparent need for re-
view of research by various types of committees; excessive formality
and documentation of informed consent, and so on (30) . Among other
things he predicts that continuation of current trends will have egregious
consequences. He expresses his alarm with the:
"...dilution and deprecation of the important by a proliferation
of the trivial. The patient, asked to sign countless releases or consents,
may respond with a blanket refusal or with a proforma signature. The
physician, immersed in a profusion of unimportant detail will loose sight
of, and respect for the important issues. Perhaps he will feel compelled
to practice defensive ethics- -no more honorable than defensive medicine.
For medical ethics, in short, trivialization is self-defeating".
Thus, he draws on the experience of observing the behavior of physicians
in reaction to their awareness of ever increasing possibilities for malpractice
litigation. Physicians now obtain many more diagnostic tests than are necessary.
This custom is referred to as the practice of defensive medicine (7, at p.
414 et seq). It is enormously expensive for all concerned. The analogy to
3-72
which he attempts to call attention—defensive ethics- -would include
highly formal negotiations for informed consent to research procedures
involving minimal involvement, risk, or discomfort to the subject.
Ingelfinger earlier expressed his view that --particularly in bio-
medical experimentation (including therapeutic innovation)--consent is
generally informed only technically but virtually never educated (31).
He argues powerfully that the individual who has assumed the sick, role
is virtually never capable of either adequate understanding or total
freedom of choice. He sees the relationship that might be created by
educated consent as described by Jonas (guora) (21) or Ramsey--a "...coven-
antal bond between consenting man and consenting man (that) makes them...
joint adventurers in medical care and progress..." as essentially un-
attainable and Utopian ideals. He suggests that it is worth striving
toward these ideals but that we should understand that they never will
be reached. He acknowledges that:
"The procedure currently approved in the United States for enlisting
human experimental subjects has one great virtue: patient-subjects are
put on notice that their management is in part at least an experiment.
The deceptions of the past are no longer tolerated. Beyond this accomplish-
ment, however, the process of obtaining "informed consent," with all its
regulations and conditions, is no more than an elaborate ritual, a device
that, when the subject is uneducated and uncomprehending, confers no more
than the semblance of propriety on human experimentation. The subject =
only real protection, the public as well as the medical profession must
recognize, depends on the conscience and compassion of the investigators
and his peers."
Benjamin Freedman agrees- -but without dismay- -that "fully informed
consent" is unattainable and suggests that striving for it is, in most
cases, undesirable (32).
3-73
He proposes that it might better serve our purposes to negotiate for
"valid consent" rather than for "fully informed consent". He concludes:
"...that valid consent entails only the imparting of that infor-
mation which the patient/subject requires in order to make a responsible
decision. This entails, I think, the possibility of a valid yet ignorant
consent."
Further: "...the informing of the patient/subject is not a funda-
mental requirement of valid consent. It is, rather, derivative from the
requirement that the consent be the expression of a responsible choice.
The two requirements which I do see as fundamental in this doctrine are
that the choice be responsible and that it be voluntary."
He sees responsibility as a dispositional characteristic which can
be defined only relatively and conditionally in the context of the tot-
ality of ones knowledge of an individual. Similarly, he claims that
"voluntarism" and reward can be evaluated only in the context of the
prospective subject's total environment.
Perhaps germane to the argument presented earlier in the present
paper--that it is a serious step for an IRB to impose a requirement for
third party scrutiny- -are some of the points Freedman makes under the
heading "The Right to Consent":
"From whence derives this right? It arises from the right which
each of us possesses to be treated as a person, and in the duty which
all of us have, to have respect for persons, to treat a person as such,
and not as an object. For this entails that our capacities for person-
hood ought to be recognized by all--these capacities including the
capacity for rational decision, and for action consequent upon rational
decision. Perhaps the worst which we may do to a man is to deny him his
humanity, for example, by classifying him as mentally incompetent when
he is, in fact, sane. It is a terrible thing to be hated or persecuted;
it is far worse to be ignored, to be notified that you "don't count"."
Freedman' s paper is strongly commended to the attention of the
reader who is unfamiliar with it; in my view, this paper analyzes in a
formal way the sorts of thinking about informed consent that ought to be
3-74
going on in IRBs. I shall conclude that in almost every situation,
the negotiations for and the documentation of informed consent ought
to be less elaborate than the possibilities presented earlier in this
paper. How much less elaborate? This should be determined on a
protocol by protocol basis (rarely on a subject by subject basis) by
an IRB which collectively is aware of the important issues in informed
consent (cf, Summary and Conclusions).
Situations requiring no consent
As noted earlier, NIH now requires full documentation of informed
consent to retain for research purposes an organ or fragment thereof
removed at either autopsy or surgery--even when the procedures them-
selves are done in accord with usual and customary medical practice.
Holder and Levine have examined the legal bases (as presented by the
Office for Protection from Research Risks) as well as the expected con-
sequences of this requirement (25). First, they find that this require-
ment is not justified by any existing case or statutory law. Secondly,
they find that this requirement is enormously wasteful of time not only
of investigators but also of the physicians who ordinarily negotiate for
informed consent to either autopsy or surgery as well as the prospective
consentors. Further, the individuals who might be asked to consent to
such procedures are those who are about to undergo surgery and the next-
of-kin of recently deceased people. It seems inappropriate to the authors
to trouble people in these circumstances with the necessity to make de-
cisions that are made to seem much more consequential than they really
3-75
are. Finally, in accord with Ingelfinger the authors conclude that what
is likely to be sacrificed in the process is the appropriate respect
given to the negotiations for informed consent in situations in which
they are ethically (and perhaps legally) appropriate. Some situations
are identified in which it might be appropriate to negotiate informed
consent to retain specimens for research purposes. These situations may
be identified by having in common two attributes:
"1) The information to be obtained by the research procedure either
will or might be linked to the name of the individual from whom the specimen
was removed. If there is no way to link the information to the name of the
individual there seems to be no possibility of putting that individual or
his next of kin at risk.
"If, and only if, the first attribute is present the proposed re-
search should be reviewed for the possible presence of the second.
"2) The proposed research may yield information having diagnostic
significance. In this regard, one should be particularly concerned if the
presence of the diagnostic information might expose the individual to lia-
bility for criminal action (eg, detection of alcohol or abuse drugs), or
might jeopardize his insurance or Workmen's Compensation status or create
the potential for civil litigation. Additionally, one should be concerned
about the possibility that the information might significantly change the
way in which the individual is perceived by himself, his family, social
group, employers, and so on (eg, detection of venereal disease)."
The preceding proposal is not be be construed as indicating that
surgical or autopsy procedures should ever be done without appropriate
authorization by those entitled to give it. However, the debates as to
whether consent for medical practice procedures should be more or less
elaborate than those for research procedures notwithstanding, in 1975,
documentation of the negotiations for consent to authorize autopsy or
surgery for non-research purposes is customarily much less formal than
3-76
it is when similar procedures are done for research purposes.
It is recognized that standard forms for consent to either surgery
or autopsy currently used in hospitals vary greatly in the language.
they contain concerning the disposition of organs and tissues. Ttous
the preceding proposal was made along with a recommendation that (25):
"...the customary practices of the hospital (should be) made ex-
plicit on the consent forms. Thus, the standard forms for consent to
autopsy might indicate- -when appropriate- -that it is customary practice...
to remove and retain some organs, tissues, and other parts as may be
deemed proper for diagnostic, research, or teaching purposes. Similarly,
the... form for. . .surgery might indicate that is is customary to deliver
the removed part to the surgical pathologist for diagnostic testing and
that in some cases it might be retained for research or teaching purposes
before it is destroyed."
There are several other commonly performed types of research activities
(eg, study of "left over" blood, urine, spinal fluid, and so on, collected
for diagnostic purposes) which may be analyzed similarly to those performed
on specimens secured at surgery or autopsy. In general, the same two at-
tributes will suffice to identify those activities for which informed con-
sent seems necessary.
Situations in which informed consent may be either not feasible or
detrimental to the interests of the research .
There are some research activities which present considerable poten-
tial harm to the subjects but, in which, any informed consent- -in fact,
the mere revelation that research is being conducted- -might defeat the
purposes of the research. Most of these activities are categorized in
the so-called "unseen observer" type of research most commonly performed
by social and behavioral scientists. The purpose of unseen observer re-
3-77
search ordinarily is to observe individual or group behavior as it
occurs in a natural environment. The potential harms of this research
include violations of privacy and confidentiality. Examples of this
type of research and the problems it presents were cited earlier (2, at
pp. 24 et seq). Problems peculiar to unseen observer research were
discussed extensively at a recent conference (33). A concise discussion
of the problems presented by unseen observer research may be found in the
A. P. A. Ethical Principles (16, at pp. 30-35).
A major unresolved problem may be expressed as a question: How
does one establish the authority to proceed with this type of research?
Estimates of potential benefits to society—the importance of the knowledge
to be gained- -are often influenced by various types of political consid-
erations. The author has proposed earlier (2, at p. 55) that one type
of research activity that might appropriately be reviewed for ethical
propriety at a national level is that in which society is the subject
and the risks to society are deemed substantial. Some types of unseen
observer research are included in this category- -even when the subject
is one individual selected and observed without his awareness.
Unseen observer research is not limited to the social sciences.
Psychiatrists, psychologists, and others commonly observe the activities
of their patients (or subjects) through a one-way glass or by recording
(video or audio tapes) their communications, interactions, and other mani-
festations of behavior. Much research involves examination of school,
employment or medical records. However, study of records and most bio-
medical research (as contrasted with some behavioral and social research)
3-78
in which the observer is unseen is not dependent for its successful com-
pletion on a lack of the subject's awareness of the observation.
Situations in which documentation of the negotiations seems un-
necessary or self-defeating.
In research activities in which it is proposed to do something that
will cause the subject only minor to modest inconvenience, that has no
potential for causing lasting harm to the subject, and that does not call
upon the subject for continuing or repeated involvement, it seems appropriate
to negotiate informed consent without the aid of written instruments. In
some cases implied consent may be quite sufficient. For example, an in-
vestigator may indicate to an individual that he wants a blood sample for
purposes of doing some basic research. If that individual offers his arm
for purposes of allowing blood to be drawn it may be said that he has granted
implied consent. In this situation it is the responsibility of the investi-
gator to inform the subject of anything that might influence his willingness
to offer blood. Thus, for example, if there is anything inherent in the
research that might meet the two attributes specified for establishing a
necessity for informed consent (cf, Situations Requiring No Consent) it is
the responsibility of the investigator to so inform the prospective subject.
On the other hand, if he merely wants a source of red cell membranes, one
would not ordinarily hold him accountable for a detailed explanation of the
purposes of his research.
In some types of questionnaire research implied consent seems ap-
propriate. The investigator may distribute copies of a questionnaire by
mail along with a covering letter indicating what protections (if appropriate)
3-79
are provided for preserving the confidentiality and privacy of res-
pondents. He might further indicate that completion of the questionnaire
and its return to the investigator implies consent. When what is being
sought are aggregate data on groups, it might, in fact, defeat the
purposes of preserving confidentiality by having subject's names on
consent forms.
In the conduct of innovative therapy--in emergency situations—it
may best serve the interests of the patient-subject to proceed without
informed consent. In some cases, the time it might take to negotiate
informed consent might delay initiation of the therapy beyond the time
during which it would be most effective. To put it most dramatically,
the patient-subject might expire during the course of the negotiations.
In situations in which the patient- subject is inaccessible owing to un-
consciousness or inebriation, it might be feasible to proceed with proxy
consent given by the next of kin. However, in true emergencies or when
the next of kin is not available it will ordinarily be considered ap-
propriate to proceed without consent. In situations in which the thera-
peutic innovation is being performed repeatedly in an institution, at
the time any particular patient-subject for whom it might be appropriate
arrives, the general plans to conduct this activity will have ordinarily
been reviewed and approved by the IRB. If the physician- investigator
anticipates that the innovative procedure might be used in emergencies
or in patient-subjects who are incapable of consenting, he should specify
this in the protocol to be reviewed by the IRB. The IRB should develop
3-80
clear understandings with the physician- investigator as to the circum-
stances in which proceeding without consent would be considered ap-
propriate. This process also assures that group judgment will be super-
imposed upon that of the physician- investigator that the particular in-
novative therapy is, in fact, the best therapy available for a particular
class of patient-subjects.
In some cases it might be considered appropriate for a physician to
proceed in an emergency situation with a therapeutic innovation that has
not been reviewed by the IRB. It might be that the particular patient-
subject is the first in whom the innovative therapy is tried in that
particular institution. Alternatively, a physician might conceive of
a novel approach to a unique problem. In such situations, the physician
should be aware that subsequently he will be held accountable to the IRB
for having attempted to meet the standards it might have imposed if con-
sulted in advance. He will also be held accountable to the patient for
an explanation of why he assumed that, had the patient been able to give
his informed consent, he would have.
Summary and Conclusions
This paper begins with an identification in its first section of
seven seperate functions that informed consent might be seen to serve.
Throughout the paper an attempt is made to relate each of the component
processes of informed consent and the devices that might be used to assure
the accomplishment of these processes to the various functions specified
at the outset.
The author concludes with Katz that informed consent remains an ill-
defined concept. In order to prepare the ground for discussing the possi-
3-81
bilitles of informed consent, there is first a presentation of how it
is defined by DHEW regulations and how it might be seen most idealisti-
cally by Katz, Jonas, and others. An attempt is made to develop a
model that might aspire to these high standards. Yet, later, (pp. 71
et seq) the conclusion is reached that it is almost never necessary to
use all of the possibilities presented earlier in the paper. Freedman's
proposal that "valid consent" might better serve our purposes than "in-
formed consent" is endorsed by this author. This conceptualization ac-
commodates the possibility of "valid yet ignorant consent".
Informed consent is a type of a contract in which one of the con-
tractors (the investigator)--as in all fiduciary relationships—is held
accountable for higher standards of responsible conduct than are most
individuals in creating commercial contracts. The negotiations resulting
in the creation of this contract are presented as having four component
processes. The investigator is assigned responsibility for seeing to it
that each of these four processes is accomplished satisfactorily.
The first is that of informing the prospective subject. Eleven es-
sential elements of information are listed in the table of contents on
the title page. Perhaps the most important is that of defining the role
of subject and how it differs from that of patient (or any other client
of a practicing professional). Particularly difficult issues to resolve
are presented under elements 5) and 6). Under element 5), the perils of
both overdis closure and underdisclosure are discussed. A most difficult
problem is that of determining which information is "material". Definitions
of materiality as developed in recent court decisions are presented. Devices
3-82
such as the "consent jury" and the "surrogate system" for making de-
cisions as to what information is material are discussed.
In section 6) a major problem is identified. How does one de-
termine the extent to which disclosure of hoped-for benefits other
than direct health benefits might be viewed as undue inducements?
Under element of information number 11) it is proposed that in
some circumstances it is appropriate to ask a prospective subject to
agree to being incompletely informed. However, lying to prospective
subjects should be avoided.
Some barriers to comprehension and devices that might be used to
overcome them are discussed. In some cases it might be appropriate to
use the "two-part consent form" to assure the investigator that the pros-
pective subject comprehends the key elements of information.
Similarly, barriers to autonomy are discussed. It is concluded
that any person who assumes the "sick role" is incapable of complete
autonomy. The definition of the sick role is deliberately extended
beyond that of patient. It is meant to include all individuals who are
somehow incapable of performing the roles and tasks for which they have
been socialized and who are expected to seek technically competent as-
sistance in "getting well". Some devices are suggested that might aid
in determinations as to whether the prospective subject is suitably
autonomous; it is acknowledged, however, that one never knows with
certainty.
Consent is not seen as the mere consummation of a contract. Rather,
it is necessary that the investigator remain sensitive to the possibility
3-83
that the subject may wish to terminate the contract at any time.
For the past 30 years there seems to have been nearly universal
agreement that negotiations for informed consent to the investigator-
subject .relationship should meet higher standards than those for the
physician-patient (or any analogous professional-client) relationship.
However, there have been some recent departures from this viewpoint.
Various opposing points of view are presented. The author concludes
that patients (or other clients of professionals) are entitled to the
same degree of thoroughness of negotiations for informed consent as
are subjects. However, the patient (client) should, in general, be
allowed more freedom than the subject to relinquish this entitlement.
As noted earlier, the subject should be made aware of how the role of
subject differs from that of client.
The possibility that supervision of the negotiations for informed
consent by consent committees might be advantageous to the participants
(and particularly to the prospective subject) is examined in some de-
tail. It is concluded that the best interests of all concerned might
be achieved more effectively and at much lower cost through the use of
third party scrutiny. There is an identification of the various in-
dividuals who might be called upon as third party scrutinizers as well
as the situations in which they are most likely to be found useful. It
is suggested that the need for third party scrutiny should be proposed
by the investigator and reviewed and approved by the IRB. Ordinarily,
the prospective subject should be allowed the option to select a third
party scrutinizer or to refuse such scrutiny. It is a very serious step
3-84
to impose on a prospective subject a requirement for third party
scrutiny; often such a requirement is tantamount to a declaration to
the prospective subject that his judgment (ability to comprehend;
ability or freedom to make choices; and so on) is questionable. How-
ever, in some cases such a requirement is necessary.
Attention is directed to existing DHEW and FDA regulations for
documentation of informed consent. It is suggested that these re-
gulations are excessively inflexible. The perils of inflexibility are
discussed in several subsequent sections. In the section entitled
Situations Requiring No Consent several perils are identified of re-
quiring fully documented informed consent in situations where it is
not necessary. Most alarming is the dilution and deprecation of the
important by a proliferation of the trivial. What is likely to be
sacrificed is the appropriate respect given to the negotiations for in-
formed consent in situations in which they are ethically (and perhaps
legally) appropriate.
In almost every research activity involving consequential risk or
inconvenience to the subject it is most appropriate to document in writing
the fact that appropriate negotiations for informed consent have been con-
ducted. The consent form can almost never be so constructed as to anti-
cipate all of any particular prospective subject's wishes to be informed.
The consent form is most effective when it is viewed by the investigator
as an instrument designed to guide him in his negotiations with the pros-
pective subject. The consent form should contain at least the minimum
amount of information and advice that should be presented during the negoti-
ations for informed consent. If any substantive new understandings are
3-85
developed in the process of the negotiation which have any bearing
on the prospective subject's willingness to participate these should
be added to the consent form signed by that particular individual.
The remainder of the issues discussed in this paper are identified
by category in the table of contents beginning with page 62. The titles
of the sections suggest the conclusions of the author.
At many points throughout this paper it is concluded that the
IRB should be assigned responsibility for determining what constitutes
the minimum standards of informed consent for each protocol it reviews.
This proposal recognizes the impossibility of defining what constitutes
adequate informed consent to research generally or even to broad cate-
gories of research. Minimum standards determined by the IRB should be
set rather high. The IRB should be aware of the perils both of under-
informing and of overinforming; when in doubt it should err somewhat
on the side of overinforming. The IRB should be so constructed as to
collectively be aware of all of the consequential issues in informed
consent. It should review proposals drafted by investigators to accomplish
the four component processes of the negotiations for informed consent; the
personnel who will be called upon to negotiate, consult, or scrutinize
the negotiations; the form and substance of the documentation of the
negotiations; and so on. Regulations should be developed that will guide
the IRB in its deliberations on the issues of informed consent to parti-
cipation in particular protocols. However, these regulations should be
sufficiently flexible to allow a well-informed IRB to meet the needs of
particular proposals to do research on particular subject populations.
3-86
These regulations should reflect an assumption of trust of the IRB by
the regulatory agency. Similarly, the IRB in its dealings with in-
dividual investigators should reflect an assumption of its trust of the
investigator. It is acknowledged that some investigators have not con-
ducted themselves in their negotiations with subjects in a way that is
conducive to developing regulations based on assumptions of trust. How-
ever, an awareness of the need to safeguard the rights and welfare of
human subjects is gradually pervading the environments in which research
is done. A changing social consciousness in the research environment is
diminishing the probability that "permissive" (34) investigators will
escape notice; in the current environment they are more likely when
noticed to be reprimanded than rewarded. Regulations based on assumptions
of trust are far less costly--financially, but even more importantly,
socially—than are those based on assumptions of distrust.
Finally, it is concluded that the nature and definition of informed
consent cannot be described definitively in the abstract. Functionally
relevant definitions can be developed only in relation to specific pro-
posals. In each case, the investigator should draft his proposal based
on his knowledge of all aspects of the research and of the prospective
subjects. The IRB should review the plans and negotiate with the in-
vestigator what has been referred to as the minimum standards for that
particular project. The investigator should then proceed to negotiate
with prospective subjects with an awareness that the plan he has agreed
to with the IRB may often have to be supplemented or modified to meet
the needs of particular subjects. Thus it is that the nature and definition
3-87
of informed consent is and ought to be continually negotiated and re-
negotiated.
3-88
References
1. Levine, R.J. : The boundaries between biomedical or behavioral
research and the accepted and routine practice of medicine. July
14, 1975; Addendum: September 24, 1975.
2. Levine, R.J. : The role of the assessment of risk-benefit criteria
in the determination of the appropriateness of research involving
human subjects. October 27, 1975.
3. Katz, J. and Capron, A.M. : Catastrophic Diseases — Who Decides What ?
Russell Sage Foundation, New York, 1975, 271 pp.
4. Katz, J.: Exper iment at ion with Human Beings . Russell Sage Foun-
dation, New York, 1972, 1159 pp.
5. DHEW Rules and regulations; Protection of human subjects; Technical
ammendments. Federal Register 40 (No. 50): 11854-11858, March 13,
1975.
6. Katz, J.: The regulation of human research--ref lections and pro-
posals. Clin. Research 21: 785-791, 1973.
7. Holder, A.R. : Medical Malpractice Law , John Wiley and Sons, New
York, 1975, 561 pp.
8. Epstein, L.C. and Lasagna, L. : Obtaining informed consent: Form
or substance, Arch . Int . Med . 123: 682-688, 1969.
9. Alfidi, R.J. : Informed consent: A study of patient reaction. J. A.M. A .
216: 1325-1329, 1971.
10. Canterbury v. Spence , 464 F 2d 772, CA DC 1972.
11. Cobbs y. Grant, 502 P 2d 1, Cal 1972.
12. Hauck, G.H. : Aspects of informed consent in research involving
human subjects: The unlikely harm. To be published in the pro-
ceedings of the conference: The Unattainable Aspects of Informed
Consent , co-convenors, Dr. Albert Jonsen and Dr. Virginia Oleson,
held in San Francisco, May 16-17, 1975.
13. DHEW, NTH: Protection of human subjects, Policies and procedures.
Federal Register 38(No. 221): 31738-31749, November 16, 1973.
3-89
14. DHEW, Office of the Secretary: Protection of human subjects,
proposed policy. Federal Register 39 (No. 165): 30648-30657,
August 23, 1974.
15. DHEW, Office of the Secretary: Protection of Human Subjects;
Fetuses, pregnant women, and in vitro fertilization. Federal
register 40(No. 154): 33526-33552, August 8, 1975.
16. American Psychological Association, Inc.: Ethical Principles in
the Conduct of Research with Human Participants. A. P. A., Inc.,
Washington, 1973, 104pp.
17. Reyes v. Wyeth Laboratories , 498 F 2d 1264, CAS, 1974.
18. Morris, R.C. : Legal problems of emergency and outpatient care.
Connecticut Med . 38: 543-548, 1974.
19. Miller, R. and Willner, H.S. : The two-part consent form: A
suggestion for promoting free and informed consent. New Eng .
J. Med. 290: 964-966, 1974.
20. Parsons, T. : Definitions of health and illness in the light of
American values and social structure, in, Patients , Physicians
and Illness , pp. 107-127, ed. by E.G. Jaco, Free Press, Glencoe,
Illinois, 1972.
21. Jonas, H. : Philosophical reflections on experimenting with human
subjects, in Exper imentat ion with Human Subjects , pp. 1-31, ed.
by P. A. Freund, George Braziller, New York, 1970.
22. Feinstein, A.R. : Clinical biostatistics : XXVI. Medical ethics
and the architecture of clinical research. Clin . Pharmacol .
Therap . 15: 316-334, 1974.
23. Fost, N.C. : A surrogate system for informed consent. J. A.M. A.
233: 800-803, 1975.
24. Position statement of the American Federation for Clinical Research
on the DHEW proposed rules on protection of human subjects. Clinical
Research 23: 53-60, 1975.
25. Holder, A.R. and Levine, R.J. : Informed consent for research on
specimens obtained at autopsy or surgery: A case study in the over-
protection of human subjects. Clinical Research 24: (in press for
February 1976 issue); preprints available upon request.
3-90
26. DHEW, Food and Drug Administration: Drugs for human use; Re-
organization and republication. Federal Register 39 (No. 62):
11684-11685 and 11712-11718, March 29, 1974.
27. Louisell, D.W. and Williams, II. : Medical Malpractice . Volume 2,
Matthew Bender, New York, 1974.
28. Gray, B.H. : Human Subjects in Medical Experimentation, John
Wiley and Sons, New York, 1975, 298 pp.
29. University of California, San Francisco, Committee on Human
Research: Guidelines , January 1, 1975, 38 pp.
30. Ingelfinger, F. J. : The unethical in medical ethics. Ann . Int .
Med . 83: 264-269, 1975.
31. Ingelfinger, F. J. : Informed (but uneducated) consent. New . Eng .
J. Med. 287: 465-465, 1972.
32. Freedman, B. : A moral theory of informed consent. The Hastings
Center Report 5(No.4): 32-39, 1975
33. The conference is that cited in reference (12). Perhaps the
Commission would wish to examine the proceedings of this con-
ference prior to publication.
34. Barber, B. , Lally, J.J., Makarushka, J.L., and Sullivan, D. :
Research on Human Subjects , Russell Sage Foundation, New York,
1973, 263 pp.
3-91
APPROPRIATE GUIDELINES FOR THE SELECTION OF HUMAN
SUBJECTS FOR PARTICIPATION IN BIOMEDICAL
AND BEHAVIORAL RESEARCH
Robert J. Levine, M.D.
February 2, 1976
The Commission is charged with the responsibility to consider: "Appropriate
guidelines for the selection of human subjects for participation in biomedical and
behavioral research." This paper is an attempt to identify and to begin to analyze
the issues that must be considered as one attempts to respond to this charge. The
following assumptions limit the scope of this paper: It is assumed that the reader
will be familiar with the three previous papers prepared by the author for the
Commission (1-3); some of the concepts developed in this paper will be incompre-
hensible without reference to the previous papers. Separate papers are being
developed for the Commission on research involving children, prisoners, the in-
stitutionalized mentally infirm, psychosurgery, and the role of the research subject
as a job. Since it is assumed that these papers will deal with guidelines for
selection of subjects in and for these categories, this paper will address only to
a limited extent problems peculiar to these populations.
The paper begins with a statement of a principle for defining the universe of
prospective subjects for any particular research proposal according to the biological
and/or social attributes necessary to test the hypothesis of the proposal. This
principle is meant to have priority over any subsequently stated guidelines. There
is next a classification of various types of research according to three separate
sets of criteria. First, a separate class is developed as one that presents mere
inconvenience to individual subjects as opposed to true risk of harm. Next there
are three categories of research which present either known or unknown risks of
physical or psychological harm to individual research subjects. A final classi-
fication is one that presents risks to groups or classes of subjects. In this
section there is a draft of some guidelines for selection of subjects which are
oriented primarily toward minimization of harm to either individuals or groups of
individuals. The remainder of the paper is concerned with selection of individual
4-1
subjects with a view toward protecting the rights and welfare of individuals.
The next section is concerned with a classification of prospective subjects.
Particular attention is devoted to identifying those who might be incapable of
protecting their own rights and welfare by virtue of being either uncomprehending,
vulnerable or dependent. In this section, further guidelines are proposed which
are designed to assist such persons in the protection of their own rights and
welfare. In some cases it is acknowledged that some persons are absolutely in-
capable of protecting their rights and welfare even with assistance; this re-
sponsibility must be assumed by others.
In the next section there is a reexamination of the various categories of
research which present either inconvenience or risk to individual subjects. There
is little further discussion of the protection of the interests of groups except
in situations in which groups might have desired dependency statuses threatened.
The purpose of this section is to examine some special cases of research presenting
either known or unknown risks of physical or psychological harm to explore the
conditions under which various sorts of subjects might be selected.
It is suggested that establishment of a suitable system of compensation for
harmed research subjects would yield a variety of salutary effects including the
facilitation of the development of satisfactory guidelines for the selection of
subjects for research presenting risk of physical or psychological harm.
The concepts of community consent and consultation, surrogate consent and
consultation, lottery systems, and reciprocal obligation, are each discussed in
detail at the first point in the paper at which, it appears it might be useful in
developing guidelines for selection of subjects (cf, Table of contents).
In general, an effort has been made to open for discussion a reasonably com-
prehensive list of issues that must be considered in developing appropriate guide-
4-2
lines for the selection of subjects. In some cases there are proposals for the
language that might be used in guidelines the Commission might wish to recommend.
Some of these are presented in language that is more definitive or more demanding
than the realities of research (or society) may be able to accommodate. More
commonly, there are presentations of general formulations that might lead to the
development of guidelines. Some difficult problems are identified which one might
wish to attempt to resolve through guidelines without proposed guidelines or general
formulations; in general, this reflects an attitude of pessimism on the part of
the author that such problems are susceptible to resolution through guidelines.
SELECTION BY REQUIREMENTS OF THE RESEARCH
Principle : The subjects of research shall have those attributes that will
permit adequate testing of the hypothesis. In most biomedical and in some behavioral
research the attributes can (and should) be stated precisely in biological terms.
In some behavioral and in most social research, the attributes can (and should) be
stated in social terms. An adequate statement of those biological and/or social
attributes that establishes a universe of prospective subjects includes criteria
for exclusion as well as inclusion.
This principle is derivative on a conclusion developed in a preceding paper
(2, at pp. 29-31):
"...There is no way to separate the issue of quality of scientific design of
research from the ethical considerations as to whether it should be done. If re-
search is badly designed it is not likely to benefit anyone. Thus, it seems in-
appropriate to put human beings at risk to develop information (or misinformation)
that cannot conceivably benefit either the individual or society."
This principle has far reaching implications in the protection of human beings.
It not only protects the rights and welfare of the individuals who are the subjects
of the research. Much harm can be done by including in the subject population in-
4-3
dividuals who are not legitimately part of that population. The most grave con-
sequences of this sort of inappropriate selection of subjects may be seen when
the purpose of the research is to develop a diagnostic test or a form of therapy
for a condition. Owing to improper selection of subjects, the efficacy of a
diagnostic test or therapeutic modality may be either overestimated or under-
estimated. A means to diagnose or treat a condition may be inappropriately
utilized by others (those who read the results of the research) who may be una-
ware that they are applying it to individuals differing in important respects
from those who were subjects of the research; for an exhaustive discussion of
the problems deriving from inappropriate or inadequate definition of the bio-
logical attributes of research subjects see Feinstein (4).
In a recent editorial commenting on the remarkably divergent results of
various surgical and other approaches to treatment of patients with unstable
angina, Pauker asserts (5):
"The variation in the prognosis of patients with unstable angina is due, in
part, to disparate definitions of acute coronary insufficiency and to differences
in the degree of coronary obstruction." Consequently, providing a patient "...with
meaningful, objective data about his prognosis is a difficult task since the
literature is replete with conflicting studies and since prognosis depends on the
availability of necessary resources and upon the skill of the physicians and
surgeons."
Our ability to define categories of people in biological terms changes with
time. Seemingly homogeneous categories tend to become heterogeneous. Thus, 40
years ago--based on the understandings of the time--it would have been appropriate
to study a form of therapy for hypertension. .Now that we know that hypertension is
not a disease but rather a manifestation of a large number of diseases such a study
would be inappropriate.
In 1975 it is necessary to develop a subject population that may be adequately
4-4
described either as having essential hypertension or as having hypertension due
to some specific cause. Within the population of those with essential hypertension
it is customary to subdivide groups for study according to certain parameters that
can be defined more or less precisely. Thus, a study may be done on the effects
of a certain form of therapy in a population described as having mild, moderate,
severe, or accelerated hypertension. These groups may be further subdivided as
having or lacking various specific complications; eg, renal, cardiac, or cerebro-
vascular. At the time of this writing there is considerable controversy as to
whether each of these groups ought to be further subdivided on the basis of whether
their plasma levels of the enzyme, renin, are high, normal, or low.
About 15 years ago it was considered appropriate to categorize all children
having certain behavioral attributes as hyperkinetic or, alternatively, having
one homogeneous category of minimal brain dysfunction (MBD) . Trials of various
forms of drug therapy and psychotherapy in various locales yielded remarkably
divergent results (6). Furthermore, the prevalence of MBD in various grammar
schools was reported as ranging from less than 1 per cent in some suburban
schools to greater than 30 per cent in some inner-city schools. With time it
became apparent that certain environmental factors could produce behavior that
was difficult to distinguish from MBD. For example, it was found that hunger-
particularly going to school without having had breakfast--could produce such
behavior in some children. Some children who were being taught by teachers who
did not share their primary languages also manifested such behavior. Correction
of these environmental factors was associated with a sharp reduction in the ap-
parent prevalence of MBD. Children in whom the diagnosis was established on the
basis of improved criteria were much more likely to respond favorably to treatment
4-5
with drugs that stimulate the central nervous system (eg, amphetamine). There
was a corresponding decrease in the frequency with which central nervous system
depressants (eg, barbiturates and major tranquillizers) were considered necessary
owing to failure of stimulant drugs. All of these revised understandings fun-
ctioned to benefit the children. For example, it must now be unusual that
hungry children are heavily sedated in order to control hyperkinetic behavior
(or--as the media of the time suggested- -drugged into submission).
In many studies it is necessary to include in the subject population in-
dividuals who are commonly called "normal controls" or "healthy volunteers".
There are no such persons. Normality and health are states of being that cannot
be proved scientifically. Thus, it is ordinarily preferred to describe these
individuals as being free of certain specific attributes of non-health or non-
normality.
In an earlier paper (1, at pp. 7-8) some examples were given from biomedical
and social research of studies in which it would have been either impossible or
self-defeating to define precisely the prospective subject population in ad-
vance. It was also pointed out that the criteria for selecting subjects may
quite properly evolve during the conduct of research as a consequence of the
early research findings.
RESEARCH CLASSIFIED BY NATURE OF RISK
In this section research will be classified in five more or less distinct
categories. These categories are defined first by the nature of the risk and
second by whether the major burden of the risk is borne by an individual sub-
ject or by the class of persons whom he represents.
First, it will be necessary to distinguish inconvenience from physical or
psychological harm. There seems to be a widespread assumption that most research
4-6
imposes on subjects a substantial risk of harm. In the view of the author
this assumption has been created by the fact that those categories of re-
search that have the greatest potentials for harm are those that have been
most widely discussed in media addressed to the public. It is equally fair
to state that there are other imbalances in the public's conception of the
relative harms and benefits of some sorts of innovative practices; eg, overly
enthusiastic reports of new therapies at times mislead the public to thinking
that they are better than they are. Some persons (prospective subjects) may
believe inappropriately that merely because a form of therapy is novel or ex-
perimental it is somehow better than those in customary use (7).
The first category of research is: 1) Research presenting mere in-
convenience to the subject. The next three categories of research are those
presenting a possibility or probability of physical or psychological harm to
individual subjects; these are: 2) Research presetting unknown risk of physical
or psychological harm; 3) Research presenting known risk of physical or psy-
chological harm; and 4) Research presenting combinations of known and unknown
risk. Most innovative therapy (practices) will be found in categories 2 and
4. The final category is entitled: 5) Risks borne by society, institutions,
or classes of people. Research conducted in this category presents risks to
groups of individuals considered as groups. The risks in this category are
not only physical and psychological, they may also be social, economic, and
legal. The research conducted in this category may be basic biological, bio-
medical, behavioral or social. Guidelines are proposed for this category that
are designed to minimize harm to the group.
4-7
1) Research presenting mere inconvenience to the subject
Inconvenience is defined (15) as a noun: "2: Something that is in-
convenient: something that gives trouble, embarrassment, or uneasiness:
something that disturbs or impedes." Inconvenient is defined as an adjective:
"2: Not convenient: giving trouble, uneasiness, or annoyance: Disadvantageous,
inopportune."
Research which presents the subject with the necessity of bearing the
burden of mere inconvenience is distinguished from that which presents the
possibility of consequential injury (determined as a function of probability
and magnitude). Research in this category employs techniques, modalities, and
interventions that have been tested sufficiently to earn the classification
(by the social device) of accepted or approved (for either research or practice
purposes) (1, at pp. 5-6). Research presenting mere inconvenience is chara-
cterized as presenting no greater risk of consequential injury to the subject
than that inherent in his particular life situation. In determining the risks
inherent in the particular life situation of a prospective subject, it is not
appropriate to consider risks irrelevant to the design or purpose of the research;
thus, for example, in considerations of the risks of most biomedical research
with a view toward categorizing proposed research as presenting mere incon-
venience (as opposed to harm) it is inappropriate to enter into the calculations
the risks of being injured in an automobile accident.
The various inconveniences of participation in research have been described
in earlier papers (1-3). In general, what is asked of a prospective subject is
that he give his time (to reside in a clinical research center, to be observed
in a physiology laboratory, to fill out a questionnaire, to be interviewed, and
so on). Often there will be a request to draw some blood or to collect urine
4-8
or feces. Certainly^ the withdrawal of venous blood may be momentarily pain-
ful and be followed by a bruise but no: lasting physical harm is done. Re-
moval of some other normal body fluids may be associated with some risk of
substantial harm; eg, brain fluid, heart blood. Ordinarily, it is possible
to do such studies in individuals who require removal of heart blood or brain
fluid for diagnostic or therapeutic purposes. In these cases the fact that
some of the fluid is removed for research purposes imposes no additional risk
or inconvenience on the subject. Removal of abnormal fluids or tissues can
only be done on those who have the diseases associated with the development
of these abnormal fluids (pleural effusions) or tissues (eg, tumors). Again,
in such persons, the interests of research and practice usually may be served
simultaneously. Removal of normal body tissues may or may not present a risk
of physical harm or inconvenience. Thus, removal of a piece of skin by standard
biopsy procedures is associated with minor discomfort lasting about 7-10 days
in suitably selected individuals. On the other hand, biopsy of the liver is
associated with a probability of approximately 0.002-0.005 of complications
sufficiently serious to require treatment (eg, blood transfusion) and of ap-
proximately 0.0002 of death. Thus,, the performance of a liver biopsy cannot
be considered an inconvenience, when this is done for research purposes it
must be considered in category 3): Research presenting known risk.
Some normal tissues are obtained as by-products of indicated surgery.
It is customary surgical practice to remove a margin of normal tissue around
the diseased tissue to assure complete removal of a tumor or of an infection.
Thus, it is ordinarily possible to get specimens of most normal tissues with-
out even causing inconvenience to the individual (for purposes of these con-
4-9
siderations I would not even call this individual a subject unless the
two attributes specified earlier exist (3, at pp. 76-77).
As was noted earlier (2, at pp. 15-19) many types of social, behavioral,
and biomedical research present the potential for social, legal, and economic
risk to the subject. An attempt to reduce the probability of economic risk
to the individual is presented subsequently (cf, Compensation). The remainder
of the risks in this category usually will not become manifest as harm unless
there are violations of confidentiality. In most cases, safeguards of confiden-
tiality can be established to the extent that all concerned can be very confident
that there will be no violations. Thus, most research dealing with confidential
information can be considered as presenting mere inconvenience. Of course, there
are exceptions; where there are barriers to preservation of confidentiality—legal
or otherwise- -the research might be considered as presenting known risk.
General formulation for guidelines : In research presenting social or legal
risks to the individual which might become manifest through violations of con-
fidentiality it is the responsibility of the investigator to inform the pros-
pective subject of the extent to which confidentiality can be assured. If the
investigator can assure absolute protection of confidentiality the research is
to be considered as presenting mere inconvenience. In such cases the investi-
gator assumes responsibility (liability) for any social or legal harm done to
the subject as a consequence of failure to maintain confidentiality.
If there are barriers to preservation of confidentiality, these should be
disclosed to the prospective subject. The prospective subject assumes the risk
of the consequences of violations of confidentiality through mechanisms of which
he has been forewarned.
4-10
2) Research presenting unknown risk of physical or psychological harm
It is often stated and indeed has been in this series of papers (1-3)
that any time we intrude a chemical, device, or other intervention upon the
body or personality of an individual there is an element of uncertainty as
to the outcome (both harms and benefits). However, some interventions have
been performed sufficiently often in conjunction with detailed observations
of their harms and benefits so that we are able to predict statistically what
the effects will be of their repeated performance in a population very similar
to that in which safety and efficacy were established. It will almost always
be more difficult to predict what their effects will be in any individual
member of that population. This sort of unknown is not the category developed
in this section.
The category defined in this section is characterized as follows: The
technique, modality, or intervention has not been performed sufficiently often
or sufficiently well in humans to permit a reasonable assessment of the pro-
bability and magnitude of harm it might produce. Some innovative therapy
(practice) is contained in this category. In cases of innovative therapy it
may also be said that the modality has not been tested sufficiently to provide
a reasonable assessment of its relative efficacy. As noted earlier (2, at p.
5a):
"...When a relatively new drug is used experimentally in the treatment
of a disease for which there exist other proved drugs, one of the risks in-
volved is that for the duration of the study the subject will be deprived of
the benefit of the proved drug. It may well be that the experimental drug
will prove to be superior to the proved drug; however, this cannot be assumed
during the course of the experiment."
When a modality that has been tested previously in one class of humans--
4-11
sufficiently well to merit the social classification of accepted or approved--
is tested for safety and/or efficacy in another class of humans, the activity
is classified in category 4)*. Research presenting combinations of known and
unknown risk.
It should be emphasized that a specific modality (drug, device, surgical
procedure, psychologic test, and so on) should rarely be classified as either
a research or practice modality. Modalities that have been sufficiently well
tested to merit the social classification of accepted or approved may still
be employed with either research or practice intent. Thus, for example the
procedure of liver biopsy was discussed in the preceding section. When a
liver biopsy is performed for purposes of diagnosing hepatitis, this activity
is practice. When the same procedure is employed to secure a specimen of
liver to determine the level of some enzyme in normal subjects (normal in
that they are free of evidence of liver disease) this is research. When a
new device is used to secure a specimen of liver from a person who is thought
to have hepatitis with the expectation that it will be possible to establish
a diagnosis based upon the availability of that liver specimen, the procedure
is classified as innovative therapy (practice).
3) Research presenting known risk of physical or psychological harm
Research considered in this category employs techniques, modalities,
and interventions that have been tested sufficiently well to predict with
reasonable certainty both the probability and magnitude of. physical or psy-
chological harm in the prospective subject" population. In the case of prophy-
lactic, diagnostic, or therapeutic maneuvers, they have been tested suffi-
ciently well to be classified as accepted or approved by the social device.
4-12
The probability and/or magnitude of physical and psychological harm is greater
than that inherent in the subject's life situation..
Beecher (8, at pp. 127-128) has provided statistics on estimated death
rates from procedures that might be done with either practice or research
intent. Among the procedures for which he provided data were cardiac cathe-
terization, liver biopsy, and general anesthesia. His estimates—prepared six
years ago--are probably high for 1976. However, they do show that such data
can be developed and I agree with him that they should be developed in order
to assess the risks of doing research. In a previous paper (2, at pp. 49-53)
there was presented an account of what constitutes an adequate description of
risks and benefits. It was pointed out that it is ordinarily possible to
predict which sorts of subjects would be most susceptible to developing known
harms of research procedures. It was further pointed out that "...potentially
irreversible harms, if detected early, may be either avoided entirely or re-
duced in magnitude. This avoidance or reduction may be accomplished by dis-
continuation of the potentially harmful research procedure. Alternatively,
minimization of a developing or nascent harm may be accomplished by therapeutic
interventions; eg, timely administration of an antidote to a 'poison'."
Deriving upon the immediately preceding discussion, three general
guidelines may be formulated; the purpose of each is to minimize harm.
The first 2 are equally applicable to research in categories 2 and 4; the
facilities requirement in the third is too stringent for most category 4
activities.
a) Guideline : Every possible effort should be made to convert known risk
of physical or psychological harm to mere inconvenience.
Illustrative examples : In a person having a tumor for which the adminis-
tration of methotrexate is indicated, studies on the metabolism or biological
effects of methotrexate may be categorized as presenting mere inconvenience. On
4-13
the other hand, administration of methotrexate in therapeutic doses to an
individual for whom it is not indicated for purposes of performing the same
studies presents a high probability of serious physical harm. See also
liver biopsy examples in preceding section.
b) Guideline : Utmost effort must be made to identify those individuals
most susceptible to harm for purposes of excluding them from the prospective
subject population.
Illustrative examples : In research designed to test the effects of
strenuous exercise in "normal" humans, one would ordinarily plan to perform
various screening tests to identify individuals with coronary artery disease
in order to exclude them. In planning research involving kidney biopsies,
utmost effort must be made to identify those individuals most susceptible to
the serious complications of the procedure; eg, those having hypertension,
bleeding disorders, solitary kidneys, and so on.
c) Guideline : Research should be conducted only by investigators who
are highly skilled in the early detection and minimization of developing or
nascent harms. The facilities immediately available to them to accomplish
these purposes should be the best known.
4) Research presenting combinations of known and unknown risk of physical
or psychological harm .
The most important type of research in this category is innovative therapy
in which the therapy has been proved safe and effective in one human population
and which is to be tested for safety and/or efficacy in another. According to
our current customs, this type of activity is most commonly conducted formally
when the innovation is a drug. However, as noted earlier, this can also be the
4-14
case when the innovation is a surgical, radiological or other procedure (1,
at pp. 13a- 16a). In general, there are two broad subsets of this category
of research. The first is that in which the modality has been proved safe
and effective for the treatment of a certain disorder in one class of humans
but which is now to be used in the therapy of the same disorder in another
class of humans. The differences in these classes of humans are determined
by such biologic attributes as age, sex, pregnancy, and altered states of
physiology that are either induced deliberately (eg, by co-administration of
another drug) or are naturally occurring (eg, coexistence of another disease).
The second broad subset is that in which a modality proved safe and effective
in the therapy of one disorder is now to be tested for efficacy in a dissimilar
disorder.
In general, in making decisions on selection of subjects for research in
these categories it should be assumed that at least the same degree of harm
will be manifest in the new subject population as was in previous populations.
Thus, the three guidelines for minimization of known harm presented in the
preceding section should be applied. It should also be understood that there
may be additional harms yet to be discovered.
The two subsets developed in this category differ most importantly on
the matter of assumptions of direct health-related benefit to the subject. In
general, it may be assumed that when the same modality is used to treat the
same disorder in persons who are biologically different from those in whom its
efficacy was established, equivalent direct health-related benefits will accrue
to the new population. This is stated as a general assumption which is not
perfectly true. Many examples can be cited, for example, of altered states
4-15
of physiology induced by diseases, prior ablation of some organ, or coadminis-
tration of some drug that will either partially or totally inactivate a drug.
However, most of these situations are either known or capable of being pre-
dicted. Undoubtedly, there will be additional surprises in
the future. However, the general assumption of equivalent benefit seems to
be justified.
In a subsequent section (cf, Research presenting combinations of known
and unknown risk--2) guidelines will be developed for selection of subjects
for this subset of research. Based on the general assumption of equivalent
benefit as well as the assumption of nearly equivalent risk (this assumption
will be developed further in the same section) it will be proposed that most
persons will be in a highly strategic position to express their personal choices
as to whether they wish to become subjects of this sort of research.
On the other hand, in cases in which a therapeutic modality is being studied
for efficacy in the treatment of a disease other than that for which it is ac-
cepted or approved, the assumption of equivalent benefit does not obtain. This
will lead to the development of a guideline for the selection of subjects for
t
this subset that is more limiting than that for the preceding subset.
5) Risks borne by society, institutions, or classes of people .
It is proposed that when it is a group that will be put at risk, informed
consent should be obtained from the group as a group. The process through
which this is accomplished will be called community consent.
Community consent is defined as informed consent negotiated with the com-
munity in which it is planned to conduct research or from which it is planned
to draw prospective subjects. The elements of informed consent to be negotiated
4-16
with the community are each of those eleven (3, at pp. 9-31) that are germane
to the proposed research. It is reemphasized that the process of creating a
condition which may be called informed consent is to be considered a negotiation;
in this case the negotiation is to be between the investigators and the community,
During the process of these negotiations it is quite possible that some aspects
of the proposal might be modified; this may be documented as was suggested for
negotiations for informed consent with individuals (3, at p. 58).
The functions of community consent are essentially the same as those cited
earlier for informed consent; these were listed in an earlier paper (3, at pp.
2-4) and have been elaborated in detail by Katz and Capron (7, at pp. 82 et seq) .
However, in some situations requiring community consent, different functions
may be given greater or lesser weight. Thus, for example, it will be seen that
in negotiating informed consent for most social policy research, it will be im-
possible to honor the first function of informed consent: "To promote in-
dividual autonomy".
The community may consent to overriding the necessity for individual in-
formed consent to several categories of research. These include social policy
research which requires arbitrary allocation of material goods and
saturation involvement of the community as well as research involving some
types of inconvenience. The community may not consent to imposing known and/or
unknown risks of physical or psychological harm on one of its members without
the consent of that member. Community consent will also be recommended for
some categories of research to be conducted in groups of subjects having de-
pendent relations to institutions when their selection is based primarily upon
their being administratively available (cf, Dependent subjects and Randomized
clinical trials). In general, in these cases the community will be offered
the option to refuse to consent to some aspects of the proposal.
4-17
Community consultation is to be distinguished from community consent.
Community consultation involves consultation with the community in which
it is planned to conduct research or from which it is planned to draw pros-
pective subjects. In general, the function of community consultation is to
assist the investigator in learning how to better conduct his negotiations
for informed consent with individuals within the community. In cases in
which individuals are asked to assume known or unknown risk of physical or
psychological harm and in which the potential benefits are expected to accrue
more to the class of persons than to the individual research subjects,
through community consultation it may be learned whether the class perceives
the potential benefits as being worthy of assuming the risks of research
even by a small proportion of their class. Another purpose of community
consultation will be to reduce the fragility of individuals who are either
highly vulnerable or highly dependent (cf, Dependent subject). Community
consultation will also be advised for establishing lottery systems (cf,
Lottery Systems).
Various procedures may be used to negotiate community consent or to
achieve community consultation; some of these will be detailed in relation
to specific categories of research for which the procedures seem most ap-
propriate.
Social policy research
Some activities in this category were discussed briefly in a preceding
paper (1, at pp. 20-25). This special case of research presenting risks to
groups will now be reviewed in greater detail.
Veatch (9) has recently surveyed some of the problems inherent in con-
4-18
ducting social policy experimentation. He defines this as "conscious and
systematic experiments designed to yield results useful in the formulation
of public policy, where this intervention involves a sample of the human
population and sometimes a control group." He explores the ethical issues
in this field in relation to established and developing ethical principles
for biomedical and behavioral research.
In social policy research, one maneuver that may be performed is the
arbitrary allocation of material goods (cash, food, shelter, education,
and so on) to one segment of the population and measuring in some formal
way the effects of such allocation on the recipients. Simultaneously, a
control group that is as similar as possible is studied to see how well they
fare without having been recipients of the material goods, Veatch contends- -
and I concur--that members of the control group are equally subjects of the
social policy experiment. Thus, in some respects this type of research is
analogous to the randomized clinical trial (RCT) (10). However, in some re-
spects it is clearly very different.
As Veatch observes (9, at p. 53) : "Social policy research may be quite
a different matter. Subjects may be assigned to treatment groups before being
contacted for enrollment. It is impossible to disguise the nature of the
treatment as is normally done in medicine. Subjects in social policy research
know in advance whether they are in the control or treatment group.... Sys-
tematically transmitting such information would eliminate some of the problems
of ac ddental discovery. Certainly the community approving agency (such as
the city council) would have to know of the different treatments. That infor-
mation should not be withheld from the subjects."
Veatch further proposes (9, at p. 47): "A critical element in gaining
legitimacy for such research is community consent openly arrived at (emphasis
added). This of course raises serious problems of determining who speaks for
the community and what would constitute a valid consent, problems that are
being faced in society generally. The most closely analogous problem in medicine
is in the establishment of pilot community-wide programs for genetic screening.
4-19
In pilot programs where efforts have been made to elicit broad-based com-
munity support (such as in the Tay Sachs screening program in Baltimore),
problems have not arisen; in other programs (including some testing for
sickle cell disease and traits), failure to obtain community consent has
generated hostility."
Further (9, at p. 55): "Responsibility for harm done in social policy
research is a complicated problem. Most social policy research will pro-
bably have beneficial consequences for most of the subjects at least in the
short run. Certain harms will be pervasive, extremely difficult to define.
There may well be psychological risks and long-term harms, but these will be
nearly impossible to measure and prove. In a school voucher program a child
may do worse in the school his parents choose than in the one he was assigned
to; in a sterilization incentive test a participant may loose part of his
family through death or divorce; in a housing experiment, readjustment pro-
blems may make a new neighborhood less attractive than the old. On both moral
and practical grounds any individual or agency undertaking social policy re-
search must face the question of responsibility for such possible consequences."
Further (9, at p. 55): "In social policy research, serious harm can come
from stopping an experiment. In housing allowance programs the subjects may
have abandoned their low-rent and possibly rent-controlled apartments; in
health insurance tests, cancelled their insurance; in educational experiments,
destroyed the older school systems; in income maintenance experiments, com-
mitted themselves to installment payments or to a standard of living from which
it would be psychologically difficult and possibly harmful to retreat."
Further (9, at p. 57): "While in medical research return of the patient
to his former condition would be clearly unacceptable, in social policy re-
search this is not unthinkable. Social policy research seems to treat subjects
much as normal volunteers would be treated in medicine. In both cases there
would seem to be a moral claim to restore the subjects' prior condition to the
extent possible. This must mean that social policy research subjects, while
they are thought to be in need of intervention to improve their lot (as sick
patients are), are not thought to be entitled to being maintained in the im-
proved condition that the researchers temporarily place them in."
Finally, (9, at p. 58): "In : social policy research the claim of the
right to participate in experiments could be a serious problem. The high
potential personal benefit from participation together with arguments of justice
and fairness produce a strong case for inclusion. Since all people of the af-
fected class can not be included initially (or the research objective would be
defeated), the selection principle must be based on some ethically defensible
argument." Unfortunately, Veatch does not specify what ethically defensible
arguments might be used to establish the selection principle.
It seems quite appropriate to appeal to principles of justice and fair-
ness to determine the right (liberty?) of persons to participate in research
4-20
particularly when they consider such participation beneficial. Subsequently,
I shall appeal to similar principles in distributing the risks and benefits
of certain categories of biomedical and behavioral research. As noted earlier,
the universe of prospective subjects for most biomedical and behavioral re-
search is determined according to biological attributes. In developing a
theory of justice as fairness, Rawls (11, at p, 104) suggests that native
assets (biological attributes) are distributed according to a "natural lottery".
An attempt will be made to demonstrate that if this concept is accepted it
might be appropriate to- distribute risks and benefits according to a man made
lottery (cf, Lottery Systems).
There are those who argue that social positions and classes are not dis-
tributed according to a natural lottery; rather, these exist as a consequence
of the deliberate acts of individuals and institutions. Thus, it may be argued
that it is less appropriate to distribute material goods within various social
classes according to a man-made lottery. However, as Rawls has observed (11,
at p. 96): "The basic structure of society. . .favors some starting places over
others in the division of the benefits of social cooperation." Entry into some
social classes may also be influenced by biological attributes. Thus, for ex-
ample, persons who are less strong or less intelligent may be more likely than
others to be on welfare. Similarly, owing to various unjust practices in our
society, certain persons may be over-represented in various social classes on
the basis of such biological attributes as race and sex.
Social policy research differs from most biomedical and behavioral re-
search in that it often depends upon saturation involvement (9) . That is to
say that it is simply not feasible to conduct most sorts of social policy re-
search without involving all or nearly all persons in the community having
4-21
the social attributes necessary to test the hypothesis. Thus, in most re-
search of this sort the concept of individual informed consent is irrelevant.
What is required in its place is community consent.
I have expressed my uncertainty as to whether it is appropriate to
distribute material goods within a population according to a lottery. Thus,
I shall suggest that it is inappropriate unless the community elects this
mechanism for distribution. I shall assume that during the process of
negotiating consent with the community, the community will be able to find
no more fair or just mechanism for arbitrary allocation of material goods
than a lottery.
Guideline : In social policy research involving the arbitrary allocation
of material goods to one segment of the population and measuring in some formal
way the effects of such allocation on the recipients, community consent is re-
quired. If there are control subjects (individuals who will be similarly
studied but who will not receive the material goods) they are to be considered
part of the community from which consent is required. The elements of in-
formation to be conveyed to the community are each of those eleven elements
described earlier (3, at pp. 9-31) that are germane to the proposed research.
Procedures for negotiating community consent
1) Ordinarily, the first approach to obtaining community consent for social
policy research should be contact with those leaders who are authorized to speak
on behalf of the community. In general they should be persons who have been
selected by the community according to a democratic process and who are held
accountable for their actions to that community. When no such persons exist
it may be appropriate to call a meeting of the community from which consent is
4-22
desired. Commonly, even when legitimate spokespersons for the community
exist, they should be encouraged to call a meeting of the community for
purposes of considering the proposed research.
2) The community should agree to rules of procedure for making de-
cisions before it proceeds to make any decisions. For example, determinations
as to what constitutes a quorum, what percentage of a quorum must vote yes to
each aspect of the decision making, and so on, must be settled before the com-
munity determines whether it wishes to participate in the research and, perhaps
most importantly, how those individuals who will receive the goods are to be
selected.
It will be assumed that the communities to be studied are selected on the
basis of their being as nearly identical as possible in their possession of the
social attributes that will permit testing of the hypothesis. Therefore, it
should be a matter of no interest to the investigators which group will serve
as controls. In some cases it might be appropriate—in the interest of opening
opportunities to the receipt of goods more widely- -to involve more communities
in the consent process than one actually intends to conduct the research in. Thus,
if one wishes to compare one community with another- -the second serving as control- -
and if one conducts the negotiations for community consent with five communities,
each of the five will be presented with a 20 per cent probability of reaping the
short-term benefits. Similarly, each will have a 20 per cent probability of
assuming the inconvenience of serving as controls.
I shall assume that the community (or collection of communities) will be un-
able to discover any mechanism for allocation of the goods that they will agree is
more just than a lottery. I shall further assume, that since the investigators will
4-23
have no interest in which communities are selected as experimental or con-
trol that they should be assigned responsibility for conducting the lottery.
During the negotiations with the community for informed consent, the
overall purpose of the research should be presented (element of information
No. 1); if the purpose is anything other than achievement of a more just
distribution of material goods within the society, I shall assume that the
community will not consent to participation. Similarly, if there exists any
a priori reason to believe that an experimental social policy will not be im-
plemented even if it proves to be safe and effective, there is no ground for
justification of the experiment.
Other research presenting risks to society , institutions , or classes of
people .
In an earlier paper the risks that might be borne by society were cate-
gorized (2, at pp. 19-29). It was observed that the Commission had already
determined that some of these risks were so substantial that some categories
of research in which these risks were presented should be reviewed for ethical
propriety at a national level. These were two categories of research on the
fetus (2, at p. 54). It was further observed that DHEW regulations required
similar review of research involving in vitro fertilization. It was further
proposed that two additional categories might be considered for similar review:
The first is that sort of activity in which society is the subject and the
risks to society are deemed substantial. The second was that in which the
risks might be borne by subsequent generations. Thus, one might include in
this latter category research and development procedures that might lead to
pollution (broadly defined) of the environment or to genetically-determined
anomalies. It was further noted that research not involving human subjects
may put society at risk (2, at p. 19). Evidence was presented that in some
4-24
cases scientists were assuming responsibility for control of such activities.
Under the rubric of social risks (2, at pp. 20-22 and 24-27) some other
types of research are described in which the risks to the group in the agg-
regate may be considerably greater than might be perceived by any individual
member of the group. Thus, one may question whether it is appropriate to
ask an individual to consent to participate in research the findings of which
might harm the group of which he is a member.
As noted earlier (2, at p. 21), there may be differences of opinion
within a group as to whether the results of research were harmful or beneficial
to the interests of that group. The determination as to whether it is ap-
propriate to proceed with research of this sort may commonly be a political
rather than a scientific decision.
Guideline : In research that presents risks to society, institutions, or
classes of people, community consent is required.
The nature of the research is nearly irrelevant; it may be basic bio-
logical, behavioral, or social research. The main criterion for establishing
this class of research is that in the view of the IRB either: a) The risks
to the group in the aggregate are greater than are likely to be perceived by
any individual member of the group; b) There is likely to be difference of
opinion within the group as to whether the results of the research might be
detrimental to the group.
In this category of research there is usually a requirement for negotiating
informed consent with each individual. This requirement may be waived only if
both of the following requirements are met: a) There is no consequential risk
4-25
of physical or psychological harm to any individual; and b) The group ac-
cording to its own established rules of procedure determines that there is
no necessity for each individual to consent.
As noted earlier (3, at p. 3) the functions of informed consent include
"To encourage self-scrutiny by the physician- investigator"; "To encourage
rational decision making"; and "To involve the public". All of these functions
may be served by negotiating for informed consent with the community in this
category of research. The investigator may find that there are greatly varied
perceptions of the purposes of the proposed research; in that case he might
wish to revise it so as to make its purpose seem more clear to the community
he wishes to study. Similarly, if the validity of the research is dependent
upon the cooperation of individual members of the community--eg, as in the
case where they might be asked to complete a questionnaire- -even if the com-
munity consents to the conduct of the research by a narrow margin (eg, 55 per
cent), the results of the study are likely to be invalid owing to sampling bias.
In this case it might be predicted that 45 per cent of the prospective subjects
would either not complete the questionnaire or deliberately distort their an-
swers to achieve their own purposes.
Undisclosed purpose research
In some studies it is necessary to inform the prospective subject that
some information is being withheld deliberately (3, at p. 30). Such studies
are most commonly found in the realm of behavioral and social research; however,
examples were also provided of biomedical research dependent upon non-disclosure
of the purpose.
4-26
Veatch has suggested (9, at p. 52):
"In those rare, special cases where knowledge of the purpose would
destroy the experiment (and only in those cases), it might be acceptable
to ask a group of mock subjects drawn from the same experimental population
if they would consent to participate in the experiment knowing its purpose.
If there is substantial agreement (say, 95 per cent), then it seems reasonable
to conclude that most real subjects would have agreed to participate even if
they had had the information that would destroy the experiment's validity."
This proposal seems to combine some of the elements of the surrogate
system proposed by Fost and the consent jury proposed by Hauck to which I
have alluded earlier (3, at pp. 20-21).
Surrogate consent
Surrogate consent is defined as informed consent negotiated with a group
of individuals that is in all important respects (biologically and/or socially)
virtually identical to the actual proposed research subject population; in this
regard they are distinguished from the surrogates to be contacted for surrogate
consultation ( infra) . The elements of informed consent to be negotiated with
the surrogates are each of those eleven that are germane to the proposed research.
Surrogate consent is said to have been achieved when a sufficient number of sur-
rogates (say, 95 per cent) state that given full information they would have
consented to become research subjects. In contrast to community consent, sur-
rogate consent can never be used to authorize research in the absence of in-
dividual or proxy informed consent. However, based upon the availability of
surrogate consent, the IRB may decide that it is permissible to proceed with
some categories of research without individual or proxy informed consent; eg,
unseen observer research.
Surrogate consultation is distinguished from surrogate consent in that
the surrogates are not virtually identical to the actual proposed subject popu-
4-27
lation. Surrogate consultation is used in cases in which community consul-
tation might be desired but the actual proposed subjects are either too un-
comprehending or too vulnerable to provide such community consultation.
Guideline : If in the view of the IRB, there is a reasonable probability
that a prospective subject population would not consent to undisclosed pur-
pose research if the purpose were disclosed in advance, the IRB should de-
velop a suitable mechanism for obtaining surrogate consent.
Unseen observer research
As noted earlier (3, at pp. 77-79) in some sorts of research it would
defeat the purpose of the research if the subjects were aware that the re-
search was being conducted. It was further pointed out that unseen observer
research is not limited to the social sciences. In certain cases the purposes
of unseen observer research would be defeated by making the subjects aware of
the research.
Guideline : If in the view of the IRB, there is a reasonable probability
that a prospective subject population would not consent to unseen observer re-
search if they were made aware of the fact of its existence, the IRB should
develop a suitable mechanism for obtaining surrogate consent.
Conflicts of interest
In many cases there may be conflicts between individuals or groups having
legitimate interests as to whether research should be conducted. The conflicts
may be between society in general on one hand, and on the other, groups of in-
dividuals who might be perceived as enemies of society. These sorts of con-
flicts are not limited to social research. Thus, for example, in studies of
the pathogenesis, prevalence, ox incidence of occupational diseases there are
4-28
often conflicts between management and labor as to whether the research
should be done. Perhaps the Commission will wish to develop guidelines
as to how such conflicts might be resolved.
There is much controversy as to whether the results of research that
has been conducted unethically should be published (9, at p. 58-59). Op-
posing points of view have been developed in detail by Levine (12) and
DeBakey (13). Such considerations are beyond the scope of this paper.
However, Makarushka (14) has suggested that even when research is done with
informed consent- -and, in all other respects is conducted ethically- -there
may be some unexpected findings that are so harmful to a group that they
should not be published. I share the author's pessimism that this view
can be translated into a guideline.
INCAPABLE SUBJECTS
Incapable is defined as an adjective (15): "1) Lacking capacity,
ability or qualification for the purpose or end in view". To develop more
fully the meaning intended of this word, incapable persons are identified
as those not meeting the definition of the adjective, capable: "4: Having
sufficient power, prowess, intelligence, resources, strength, or other needed
attributes to perform or accomplish".
Thus, incapable subjects will be defined as those lacking sufficient
power, prowess, intelligence, or resources to either perform or accomplish:
a) participation in the research, or b) the protection of their own rights
and welfare.
Those persons who might be incapable of participation in research by
virtue of lacking the necessary biological or social attributes have been
4-29
discussed earlier. It should also be noted that — particularly when co-
operation on the part of the subject is required- -some individuals having
the necessary attributes to test the hypothesis may be too uncomprehending
or feeble to do what is expected of them. The primary focus of this section
is to discuss the problems associated with selection of research subjects who
are incapable of protection of their own rights and welfare. These individuals
will be further classified as follows: 1) Uncomprehending; 2) Vulnerable;
and 3) Dependent.
It should be understood that each of us when measured against the highest
standards of capability will be found relatively uncomprehending, vulnerable,
or dependent upon someone or something. Each of these categories should be
viewed as consisting of a spectrum ranging from slightly to absolutely in-
capable. The following discussion will be concerned only with those sorts of
individuals whose capabilities might be reduced to the extent that we might
wish to treat them differently from most other individuals as we are considering
the selection of research subjects. Also, in each category it will be recognized
that incapabilities may be transient, prolonged, permanent, or recurrent. These
factors will influence the assessment of each incapability as it relates to the
selection of subjects.
1) Uncomprehending subjects
Some barriers to comprehension were discussed earlier (3, at pp. 31-33);
some devices that might be used to overcome some of these barriers have also
been discussed. Among the problems that may be presented by inclusion of the
uncomprehending as research subjects are: Consent may be invalid by virtue of
this incapacity. In research activities requiring cooperation, the subjects may
4-30
not understand how they are to cooperate and, thus, either defeat the pur-
poses of the study or contribute to their own harm.
Persons having prolonged or, perhaps, permanent incapacity to comprehend
include the mentally retarded, the uneducated, and the senile. Language
barriers may sometimes be overcome with the aid of translators.
Persons with various psychological disturbances may be incapable of
comprehension transiently, for prolonged periods, permanently or intermit-
tently. In some persons this incapacity may be absolute. Some very with-
drawn schizophrenics might comprehend but give no evidence that they have.
Some intelligent individuals, during obsessional states, may be so preoccupied
with their obsessions that they have no energy or interest available to com-
prehend anything else.
Persons who are inebriated may have incapacities to comprehend ranging
from barely perceptible to absolute. In some cases the inebriated person
may insist that he is better able to comprehend than another person has:told
him he is. Inebriation is ordinarily intermittent. Most often, a person who
can reasonably assume that he will be inebriated at some future time is
capable of comprehending plans to involve him in research during his inebriation.
It may be, for example, that his inebriation will be induced by administration
of some drug for therapeutic purposes; eg, a narcotic, barbiturate, and so on.
Under some circumstances it might be appropriate for a person to consent in
anticipation of inebriation to being treated in a certain way when he is in-
ebriated even though while inebriated he might object to being treated in that
way.
4-31
Unconscious persons are absolutely uncomprehending. Unconsciousness
may be intermittent and reasonably predictable as in individuals with grand
mal epilepsy. Such individuals are ordinarily capable of valid consent in
anticipation of their next unconscious period. Similarly, a plan may be
developed between a physician and a patient to produce unconsciousness--
eg, through general anesthesia--at some future time. This individual can
consent in advance to various research procedures that might be done during
the period of unconsciousness.
In summary, the uncomprehending subject is classified as incapable
largely by virtue of being unable to comprehend the information necessary
to provide valid consent. The following are general formulations for the
development of guidelines for the selection of uncomprehending subjects.
a) In -those who are intermittently or predictably inebriated or un-
conscious, consent can usually be negotiated in advance. If there is some
cause to suspect that one's judgment might change while inebriated, it should
be negotiated in advance whether the change in judgment is to be honored.
More specifically, the person who expects to become inebriated may consent
to have his judgment while not inebriated take priority over any expressions
to the contrary while inebriated.
b) In developing guidelines to participate in research that presents
mere inconvenience as opposed to risk of physical or psychological harm,
standards for comprehension should not be set too high. If they are, it
will tend to exclude inappropriately some persons from assuming the role of
subject as a job. In general, it should be assumed that a person is capable
of suitable comprehension unless this is demonstrably not the case. When in
4-32
doubt, comprehension can be assured with the aid of third party scrutiny
(3, at pp. 46-52) or the 2-part consent form (3, at p. 33).
c) In developing guidelines to participate in research that presents
known and/or unknown risk of physical or psychological harm, standards for
comprehension should be set rather high. To the extent that the benefits
of such research will accrue more to the class of persons represented by the
subject than directly to the uncomprehending subject there should be an in-
creasing presumption of the necessity for third party scrutiny or the 2-part
consent form.
d) In cases in which the intervention is categorized as innovative
therapy and in which the physician-investigator and the IRB agree that this
is as good as the best alternative therapy (innovative or otherwise) for a
well-defined category of prospective subjects, if neither direct nor proxy
consent is possible, consent may be presumed. In rare cases it may be ap-
propriate to proceed without first having consulted the IRB (1, at pp. 17a-
18a). However, in such situations the physician (or other health care pro-
fessional) should be aware that subsequently he is likely to be held ac-
countable for his actions to both the subject and the IRB.
e) The truly uncomprehending subject is, in general, not a suitable
prospective subject for research unless the only individuals that possess
the biologic attributes necessary to test the hypothesis are necessarily
also uncomprehending. In such cases it will be necessary to have suitable
proxy consent. In cases in which there might be difficult choices between
reasonable alternatives and in cases in which there is known and/or unknown
4-33
risk assumed by individuals with a primary purpose of benefiting their class
rather than with an expectation of direct health-related benefit, it may be
appropriate to establish a mechanism for community consultation with proxies;
this would serve to reduce the fragility of the individual proxy in negotiating
informed consent with the investigator (cf, Dependent subjects). In some cases
it might be appropriate to establish a mechanism for surrogate consultation.
Thus, for example, a proposal to perform research on a group of uncomprehending
schizophrenics might first be presented to a group of individuals who had once
been uncomprehending schizophrenics. They might be asked if they would wish
to have such research performed on them at a time when they were uncomprehending.
The information obtained from such a group might be of value to the group as-
sembled for purposes of community proxy consultation.
2) Vulnerable subjects
Vulnerable is defined as an adjective (15): "1: Capable of being wounded:
Defenseless against injury". It should be quite clear that many of the in-
dividuals that have been classified as uncomprehending are also vulnerable by
virtue of that incapacity. For example, who could be more vulnerable than the
unconscious person? Similarly, many individuals who are vulnerable by virtue
of being sick are also dependent.
Vulnerable individuals will be categorized further into two groups each
deriving its vulnerability from different origins; these are: A) The sick;
and B ) the impoverished. A third category will be developed as the potentially
vulnerable: C) Minorities.
A) The sick . The sick role and the barriers it presents to autonomy have
4-34
been discussed earlier (3, at pp. 34-39). A preliminary description of the
relative vulnerability of various persons who are in the sick role has been
presented earlier (2, at pp. 51-52). The definition of the sick role was
written in a time when most definitions of diseases were prepared by physicians
and most spokespersons for the public did not challenge these definitions. At
that time all concerned were aware that some persons assumed the sick role more
or less illegitimately. Those that did so most illegitimately were branded
malingerers and those whose assumption of the sick role was controversial were
branded hypochondriacs. Some individuals were assigned the designation sick
against their wills; this mainly occurred in disputes over whether various
sorts of personal behavior might be considered sickness as opposed to personal
preference.
Now it is clear that over the past 20 years, society- -sometimes with
and at others without the collaboration or concurrence of health professionals--
has begun to extend definitions of the sick role. Now we find that abusers of
alcohol and various other drugs that affect cognitive function are assigned
the sick role. Persons with various difficulties in learning how to read
are treated as if they were sick. In some circumstances, voyeurism, pros-
titution, gambling, violence, and other such behaviors are considered sick-
nesses. A particular problem is presented when some of these behaviors
that are seen as sicknesses are also seen as crimes. Thus, for example, a
prostitute may be offered the "option" of either going to prison or enrolling
in a "therapeutic" program.
One peculiarly anomalous situation is that of the homosexual in 1976.
In some jurisdictions homosexual behavior is considered criminal. Recently,
4-35
the American Psychiatric Association voted that homosexuality was not an
illness. Thus one wonders whether a homosexual can be offered the "option"
of enrollment in a therapeutic program as an alternative to imprisonment.
Also, one wonders whether a homosexual who perceives his own homosexuality
as an illness of which he would rather be rid can appeal to a psychiatrist
or a psychologist for therapy.
For the present purposes, sickness will be defined as a state as per-
ceived and defined by the individual assuming the sick role. Further, it
will be defined as a state which the individual would abandon- -with the aid
of technically competent help--in favor of a condition that he would perceive
as healthy or at least more healthy than he is.
Persons with short term illnesses from which they might be expected to
recover completely without the aid of health professionals will not be con-
sidered vulnerable. Persons with illnesses from which they might be able
to recover completely but only with the aid of standard (non-experimental)
technically competent help will not be considered particularly vulnerable;
this statement is based on the assumption that such help will never be with-
held as punishment for refusal to become a research subject.
Persons having prolonged chronic illnesses which are refractory to
standard therapeutic modalities may be seen as seriously vulnerable. This
becomes increasingly true when such people perceive themselves as "desperate"
and willing to take "any risk" for even a remote possibility of relief. In
this category are some infertile persons or couples who "desperately" want
a child; some persons with chronic, painful, and disabling disorders such
as rheumatoid arthritis; some obese persons who cannot loose weight following
4-36
standard procedures; some persons with severe depression, obsessive-compulsive
disorders; and so on. Depressed persons and others who question their self-
worth may be especially vulnerable to inappropriate inducement by offering
benefits other than those directly related to their health (2, at pp. 39-41).
Perhaps the most vulnerable group among those who have assumed the sick
role are some persons who believe (correctly or incorrectly) that their own
death is imminent. This will be an especially difficult group for which to
develop guidelines.
It has been proposed by Gaylin (16) that a new class of subject called
the " neomort " might be developed. A neomort is an individual who has been de-
clared dead by virtue of current criteria for establishing brain death whose
other bodily functions are maintained more or less indefinitely with the aid
of various devices. He suggests that we now tolerate such activity to a
limited extent. Persons who have been declared dead by brain death criteria
and who are to be donors of an organ now have their biologic functions main-
tained by perfusion of various fluids and nutrients and with the aid of "life-
sustaining" devices until the transplant recipient is prepared to receive the
organ. A person might volunteer for neomort status in much the same way as
one now consents to posthumous donation of an organ under the Uniform Anatomi-
cal Gift Act. Presuming their lack of sentience, these persons might be main-
tained indefinitely for a variety of biological research procedures that for
various reasons one might not wish to perform on sentient beings. Addition-
ally, organs and tissues might be harvested as necessary for transplant pur-
poses.
I shall not comment further on neomorts; perhaps the Commission will
wish to develop guidelines in this regard.
4-37
General considerations for the development of guidelines to protect
subjects who are vulnerable by virtue of being sick:
a) Persons having prolonged chronic illnesses which are refractory
to standard therapeutic modalities are particularly vulnerable to taking
risks for even remote possibilities of relief when they perceive themselves
as "desperate". In the past, such persons were particularly vulnerable to
exploitation by individuals offering "miracle-cures" such as copper bracelets,
rainbow pills, Krebiozin, and so on. Such persons are most likely to be pro-
tected by requirements that all innovative therapies are to be: "conducted
according to the highest standards of the relevant scientific discipline (1,
at p. 16a)." This presumes that they will first be reviewed by an IRB which
will have sufficient expertise to review the proposed innovative therapy for
scientific merit.
b) Persons who question their self-worth may be especially vulnerable
to inappropriate inducement by offering benefits other than those directly
related to their health. Thus, it seems appropriate that the guidelines
reflect a need for caution not to exploit such persons. On the other hand,
advice given to those responsible for selection of subjects should not be
too rigid. Some persons by virtue of being excluded might have their senses
of self-esteem diminished further.
c) The most difficult group withwhich to deal will be those who perceive
themselves as dying. Some of these persons are particularly susceptible to
loneliness; inappropriate exclusion of such persons from "normal human acti-
vities" might intensify their feelings of loneliness. Thus, it would be clearly
4-38
inappropriate to develop rigid guidelines excluding dying persons from re-
search characterized by mere inconvenience.
On the other hand, it is frequently alleged that in the past, dying
subjects were commonly exposed to high risk research not conducted for
their benefit because it was impossible to cause any lasting harm. Examples
of exposure of dying persons to high risk research not done for their benefit
have been assembled by Katz (17, at pp. 1054-1062).
In general, research on dying persons which presents known and/or un-
known risk should be limited to that in which the dying person, --or at least
that class of persons which he represents—is likely to receive direct health-
related benefits. Experiments designed to use the dying person merely be-
cause there is no further way of causing him damage—particularly when these
are done without meticulous negotiations for informed consent — are an affront
to the dignity of such persons. It is my undocumented view that such ex-
perimentation is now rare. On the other hand, Spiro has presented the equally
undocumented view (18) that the person with cancer is often viewed as de-
humanized, "and too often studies that would not be done on the healthy are
projected for him simply because he is going to die sooner than the rest."
On the other hand, guidelines should not be so rigid as to proscribe al-
truistic acts done with meticulously negotiated informed consent such as don-
ation of an organ or tissue according to the provisions of the Uniform Anatomi-
cal Gift Act.
d) Those who are vulnerable by virtue of having assumed the sick role
are, in general, the most suitable subjects for most biomedical research. That
is, they have the biological attributes necessary to test the hypotheses of most
4-39
biomedical research. To Che extent that the research is done with the in-
tent of bringing direct health related benefit to the subject this seems
to present little problem; exceptions may be seen in cases in which sick
persons might have dependent relationships jeopardized (cf, Dependent sub-
jects).
A greater problem is presented when sick persons are asked to assume
known risk with the expectation that the benefits will redound more to the
class of persons whom they represent than to themselves. In estimating the
potential benefits of research to a class of persons, the investigator is
likely to err, if at all, on the side of overestimating. The individual
sick person tends not to be strategically placed either to disagree with
the investigator's estimate or to refuse an appeal to do good for others.
Thus, it will be proposed that in research presenting known risk or in some
categories of research presenting unknown risk (where there is some cause
to presume that the unknown risk might be substantial) and the purpose of
the research is to develop information that is more likely to benefit the
class than the individual, community consultation is required. The class
of persons who will receive the benefits is better situated than either the
investigator or the individual prospective subject to decide whether the
benefits are sufficiently great that they would wish to expose even a small
number of their class to the known or unknown risks. In addition, at a
meeting of prospective subjects the fragility of the individual should be
reduced. In some cases it may be appropriate to establish lottery systems
to determine who will be the subjects of such research (cf, Lottery systems)
4-40
In some cases the proposed subject population may be so fragile that
surrogate consultation may be required (cf, Known risks--2).
e) In some cases there will be competition for scarce innovative therapies;
it will be proposed that the most fair way to determine who shall be the
recipients among those having equal medical needs will be a lottery.
f) Commonly, those who are vulnerable by virtue of having assumed the
sick role are also dependent upon either an institution or a professional. Thus,
some additional problems in selection of vulnerable subjects are discussed in
detail in the section on dependent subjects.
B) The impoverished : To the extent that a person lacks the funds to
purchase what he considers the necessities of existence, his freedom to re-
fuse to consent to do anything that might improve his purchasing power may be
limited. Within a population sharing comparable incomes various individuals
may assign different priorities to television sets, food, heat, clothing, and
so on. Some persons with what appear to be relatively high incomes do ex-
traordinary things to earn sufficient money to purchase what they consider
the necessities of existence; eg, a third family automobile or membership in
a country club. It is not the purpose of this paper to analyze the responsi-
bility of individuals to budget their resources appropriately. However, it
may be recalled that some regulations and policies on disbursement of public
funds for health or welfare seem to be developed to prevent what policy-makers
perceive as irresponsible budgeting; eg, funds may be provided as food- stamps
which are not legitimately negotiable for anything other than food. Similarly,
to become eligible for some types of publicly funded health care one must dis-
4-41
pose of any real property above a specified level.
For purposes of identifying a class of people who might be considered
vulnerable by virtue of economic impoverishment, impoverishment is defined
as a condition in which a person considers it necessary to take extraordinary
risks to secure money or other economic benefits that will enable him to
purchase what he considers to be the necessities of life. Further, his
willingness to take extraordinary risks is based upon his belief that he
cannot secure a sufficient amount of money by ordinary means. This category
intersects but is not identical with that of dependent by virtue of being on
welfare.
The Commission may wish to consider the following guideline to protect
the impoverished:' Economic inducements to participate in research should be
determinedby the amount of inconvenience to be imposed on the subject and
should not be based on calculations of known or unknown risk. As a corollary
to this guideline it might be stated that research subjects should not be
paid to assume risk; they should rather be compensated for injury.
This guideline, if adopted, would represent a departure from the tradition
in our society that persons are commonly paid high salaries to assume large
risks. Therefore, in order to justify adoption of this guideline it will be
necessary to provide justification for this departure from tradition.
It is customary to conduct most research activities without cost to the
subject. At times, an impoverished person may find himself faced with the
necessity to assume a large financial burden to purchase needed medical care
for himself or for a dependent. He may find that- if he or his dependent is
a suitable prospective subject for research, the costs of the needed medical
care may be largely underwritten by the agency sponsoring the research. Thus,
4-42
the "choice" between being a patient or a subject may be based primarily on
financial considerations. I doubt that this dilemma can be entirely resolved
through the development of guidelines. In particular, when choosing between
an innovative therapy and a customary medical procedure, if the innovative
therapy is provided free, some persons will choose the innovation largely
on that basis. A guideline might be developed that would require the subject
to finance innovative therapy much in the same way the patient finances medical
care. This seems inappropriate to me; particularly, in the early stages of
testing the innovative therapy- -when the risks and potential benefits are
largely unknown and when the testing is done for the benefit not only of
the subject but also of like persons--it seems inappropriate to call upon the
subject to assume not only the direct health related risks but also the
financial burdens.
It may be possible to minimize the possibility of some undue economic
inducement through the development of a guideline. This guideline would not
resolve the issues of choice between innovative therapies and customary medical
practice. It assumes that it is appropriate to provide economic inducement
to assume inconvenience.
General formulation of a guideline : In calculating the costs of health
services to an individual who will simultaneously play the roles of subject
and patient, it should be calculated how much his medical care would have
cost had he not agreed to play the role of subject. This amount should be
paid by the patient. The additional costs incurred as a consequence of agree-
ment to play the role of subject should be paid by the investigator or spon-
soring agency.
4-43
C) Minority groups . Individuals who are members of minority groups
(as determined by race, sex, ethnicity, and so on) are not to be considered
particularly vulnerable a priori. Because the group may be the object of
discriminatory societal customs, an inordinantly high percentage of the
group may be classified as uncomprehending, vulnerable, or dependent by
virtue of other criteria. Such persons should be treated accordingly.
The interests of members of minority groups are probably best protected
by representation on the IRB.
3) Dependent subjects
Dependent is defined as an adjective (15): "2b: Unable to exist,
sustain oneself, or act suitably or normally without the assistance or
direction of another or others." "c : Connected in a subordinate relation-
ship... d: Lacking the necessary means of support and receiving aid from
others".
Most persons are dependent upon some other persons or institutions for
most of their lives. For purposes of this discussion we shall focus on de-
pendent relationships thatpresent either of two potentials. The first is
that by virtue of the relationship, the dependent individual is administratively
more available to the researcher to be selected as a subject than are other in-
dividuals not having the same dependent status. The second is that in which
the dependent individual might fear that he might forfeit either in part or
in whole his dependent status by virtue of refusing to become a subject.
The institutions upon which an individual might become dependent and
which present these two potential problems include health delivery systems,
educational institutions, welfare agencies, places of employment, and so on.
4-44
Within such institutions the problems become more serious when the in-
dividuals have no reasonable alternatives. For example, a person who
is dependent for his health care on Veterans Administration Hospitals will
be viewed as more dependent than one who has his choice between several
private hospitals in the same city. Clearly, it is common that individuals
develop more intense dependent relationships with certain individuals within
these institutions. Thus, for example, Spiro (18) suggests that a physician
having a close relationship with a patient can ordinarily persuade that patient
to consent to nearly anything.
The author accepts the notion that dependent persons are particularly
liable to feel that their dependent relationships may be jeopardized if
they refuse to consent to become research subjects (2, at pp. 14-15); this
is particularly true when the professional with whom they are negotiating
informed consent is the one with whom they have established an intense de-
pendent relationship and when that professional has a vested interest in
having the individual assume the role of subject. Recent discussions by the
Commission suggest that many of them share this assumption. For example,
in discussions of prisoners and patients as prospective research subjects
there is often reference to the "fragility of the individual" who is either
vulnerable (by virtue of having assumed the sick role) or dependent in
negotiating with the more powerful professional. The previously described
process of community consultation may be used to reduce the fragility of the
prospective dependent research subject in relation to the investigator (cf,
Lottery systems).
4-45
General considerations for the development of guidelines for the
selection of dependent subjects :
In general, it is appropriate to draw upon the administratively avail-
able to assume inconvenience. However, in research presenting risk of
physical or psychological harm, as the degree of risk increases there
should be a decreasing assumption of the validity of drawing upon the ad-
ministratively available. In situations in which known risk is presented
in order to develop basic information that will be of presumed benefit to
the class represented by the administratively available, there should be a
decreasing assumption of the appropriateness of drawing upon the administra-
tively available without consulting them as a group. Thus, in some cases it
will be necessary to seek community consultation (cf, Lottery systems) and in
others, community consent (cf, Randomized clinical trial).
In general, it is not appropriate to threaten individuals with loss of
dependent status owing to unwillingness to assume known and/or unknown risks
of physical or psychological harm. Some rare exceptions to this generality
will be stated. For example, there are some cases in which research presenting
known and/or unknown risk is necessary to establish a category of employment;
in such cases it may be appropriate to offer such employment contingent upon
an individual's willingness to be a subject of such research (cf, Research
presenting unknown risk of physical or psychological harm--2).
Most innovative therapies are viewed as presenting either unknown or com-
binations of known and unknown risk of physical or psychological harm. In
general, it is assumed that most persons can choose to receive an innovative
therapy. However, the choice to receive an innovative therapy imposes upon
the individual the reciprocal obligation of assuming the inconvenience of tests
4-46
necessary to prove its safety and/or efficacy. In cases in which the in-
novative therapy is scarce, refusal to participate in tests necessary to
prove its safety and/or efficacy may justify exclusion of that individual;
examples of such research have been presented earlier (3, at pp. 28-30).
However, a distinction must be made between depriving a person of a single
modality (innovative therapy) on the one hand and, on the other, of an en-
tire dependent relationship. It is not appropriate to terminate a pro-
fessional-client relationship because the client refuses to be the subject
of research- -including innovative therapy.
Categories of dependent persons that will not be discussed further here
are children and prosoners. There will be some additional discussion of
children in the section entitled research presenting combinations of known
and unknown risk of physical or psychological harm--2. However, in general,
children and prisoners cannot be considered as can the sorts of dependent
persons discussed in this paper. The issues involved are entirely different
in kind.
COMPENSATION
In an earlier paper (2, at pp. 27-29), mention was made of the need to
develop adequate plans for compensation of subjects who are harmed by research.
Attention was further called to the Secretary's (DHEW) Task Force on the Com-
pensation of Injured Research Subjects along with a suggestion that the Com-
mission might wish to establish formal communications with this group. Several
approaches to insuring research subjects were also reviewed. It is the opinion
of the author that it will be most difficult to develop appropriate guidelines
4-47
for the selection of subjects for research having either unknown or known
risk of physical or psychological harm without first having made plans for
adequate compensation of those who are harmed. Subsequently stated guidelines
are based on the assumption that appropriate mechanisms for compensation of
harmed subjects will be developed.
A recent article by Adams and Shea-Stonum reviews the various approaches
that might be used to compensate harmed subjects (19). This article calls
attention to the fact that early drafts of the legislation leading to PL 93-348
would have charged the Commission with the task of developing "a mechanism for
the compensation of individuals and their families for injuries or death proxi-
mately caused by participation of such individual in a biomedical or behavioral
program." However, as the tenure of the Commission was reduced from five to
two years and the authority from regulatory to advisory, the language of this
charge to the Commission evolved accordingly. Thus, the Commission is charged
to "make recommendations to the secretary. . .concerning any other matter (in
addition to administrative procedures necessary to implement ethical guidelines
in conducting research) pertaining to the protection of human subjects of bio-
medical and behavioral research.". Thus it appears, that the legislative in-
tent was that the Commission might examine the issue of compensation and make
recommendations accordingly.
Adams and Shea-Stonum identify three types of situations in which subjects
might be injured (19, at pp. 614-616):
"First, the common negligence of the researcher-physician may cause injury
to the volunteer. Secondly, injury may arise in an experiment that should not
have been conducted in the first place. Experiments that failed to meet the
prevailing standards for approving the use of human volunteers or that lack ade-
4-48
quate consent of the subject fall within the second category. In such
situations the injury complained of may or may not related to a defect in
the execution of the experiment. Finally, injury to the volunteer may
occur notwithstanding the absence of negligence, the appropriate approval
of the experiment, and the subject's valid consent. Such unavoidable in-
juries are the specific untoward results that have given society pause in
accepting human experimentation."
In the first two types of situations there are available appropriate
common law remedies; ie, actions for malpractice. It is the third type of
situation that requires a new development.
It seems clear (19, at p. 623) that in the medical practice context,
the patient assumes the risk of any harm to which he has consented. "In the
experimental situation, however, the unknown risk is the greatest source of
potential injury. Consequently, a subject who consents to experimentation
with the specific knowledge that injury may result from an unknown risk and
who suffers harm from such a cause would not be afforded a remedy under the
present common law concept of consent."
After reviewing various legal approaches to providing economic remedies
for harmed research subjects, the authors conclude that the most appropriate
might be to develop a type of compensation fund. This fund would be estab-
lished and maintained by the federal government for all research including
that not sponsored by the federal government. In theory it is developed as
a modified workmen's compensation plan. Like workmen's compensation, subject's
compensation would be largely a no- fault system. Unlike workmen's compensation,
there would be the possibility of resorting to common law remedies if the in-
vestigator or sponsoring agency were at fault. In particular, if the investi-
gator proceeded negligently or if the research were done without appropriate
1KB sanction, either the harmed subject or the compensation fund could pro-
ceed against the investigator or sponsoring agency.
The premise on which this latter proposal is based is contained in the
following paragraph (19, at p. 639):
"In the field of medical experimentation, fault is not the two-sided coin
4-49
that it is in industrial accidents. Injuries will rarely, if ever, result
from the subject's negligence or fault. Therefore, the reasons for limiting
coverage in workmen's compensation to injuries affecting the ability to work
and for denying recovery for pain and suffering are not operative in the re-
searcher-subject relationship. Furthermore, since subject fault is not an
issue, the researchers have no basis for demanding that their fault also be
ignored. In making reference to the workmen's compensation model, we must
not lose sight of the policy and politics that caused it to preempt the
control system of the common law."
Although I must disagree with the premises for this argument- -there are
many ways in which subjects might contribute to their own harm (eg 3 at pp.
16-17)--I must agree with the conclusion that harmed subjects might seek
common law remedies for harm occurring as a consequence of investigator
negligence.
This proposal also includes a system of indemnification of the com-
pensation fund by the sponsor of the research which is based upon calcu-
lations of the probability and magnitude of both harm and benefit and also
considerations of whether the benefits are expected to be enjoyed by in-
dividual subjects or by others. It is proposed that this system—particularly
the component having to do with judgments of the amount of indemnification- -
would yield salutary effects to the entire process of determining the ap-
propriateness of any particular research proposal. It would at least clarify
the costs of conducting any particular proposal in economic terms. In my
view, it would also facilitate the development of appropriate guidelines for
selection of subjects.
I do not think the system offered in this article is ideal. However,
it does provide a point of departure for considering a suitable system as
well as a good account of the advantages that might be afforded by one. One
problem that is not addressed is that most workmen' s compensation funds are
4-50
based on the premise that all persons who might be compensated hold jobs
for which they receive predictable amounts of income in the institution in
which they might be harmed. Thus, for example, it is relatively easy to
calculate how much income is lost by the workman who is incapacitated for
2 months from performing a job that yields a salary of $4.00 per hour. In
order to apply the concepts of workmen's compensation to compensation for
harmed subjects, it must be determined whether the compensation would be
relative to the subject's expectations in other jobs he might hold. Thus,
it might be determined that the individual whose only income is welfare might
be entitled to no compensation for prolonged disability. This would appear
to be a peak achievement in inequitable compensation. On the other hand, a
system that provided equal compensation for all subjects harmed in a similar
fashion would cause society to impose known and unknown risks largely on the
economically disadvantaged.
I should also like to take exception to the notion that the compensation
fund must be established and maintained by the federal government. It would
be more appropriate to have compensation funds developed as they now are for
workmen's compensation by various employers and sponsors in accord with ap-
propriate state and federal regulation. I see no cause to distribute the
"costs" of research uniformly throughout society unless the profits will be
similarly distributed.
LOTTERY SYSTEMS
In recent years there have been several proposals that it might be most
appropriate to select subjects for certain types of research according to
lottery systems. Some authors have focused on the problem of distributing
4-51
the risks of research in accord with principles of justice as fairness (11).
Thus, Fried (10, at p. 64 et seq) , in discussing the special case of research
called the randomized clinical trial (RCT) accepts the notion that in in-
novative therapy the appropriate prospective subject population is developed
according to a natural lottery. He then proceeds to argue against those who
say it is appropriate to determine who among the prospective subject popu-
lation will receive which treatment (of two or more that are to be compared)
according to a lottery. I must agree with his argument. The real problem
is to determine fair systems for selection of subjects among whom competing
modes of therapy will be tested.
Capron (20, at p. 147), in discussing another special case--drug trials
in children--has suggested that a lottery might be used to determine who will
be the subjects of research. Having accepted the need for phase I drug testing
in children, he suggests:
"If we are, in fact, speaking of a definable group and if all members of
the group have a roughly equal chance of being afflicted with the diseases
which are to be treated in the group, then consideration should be given to
selecting experimental subjects from the group on a random basis."
The system suggested by Capron is rather like the selective service
system; it is proposed as a system that would obviate the requirement for
informed consent.
Others have focused on the utility of lottery systems to assure equal
access on the part of all persons to therapies (innovative and otherwise)
for which demand exceeds supply. Thus, for example, the Artificial Heart
Assessment Panel of the National Heart and Lung Institute (21, at pp. 197-198)
concludes :
4-52
"(5) In the event artificial heart resources are in scarce supply, de-
cisions as to the selection of candidates for implantation of the artificial
heart should be made by physicians and medical institutions on the basis of
medical criteria. If the pool of patients with equal medical needs exceeds
supply, procedures should be devised for some form of random selection. Social
worth criteria should not be used, and every effort should be exerted to mini-
mize the possibility that social worth may implicitly be taken into account."
Outka (22) has analyzed in detail the issues of equal assess to health
care in relation to principles of social justice. He concludes most strongly
that considerations of merit, desert, societal contribution, and so on, have
no place in making decisions on the allocation of the goods of health care.
The most important principle he develops is: "To each according to his es-
sential needs." Essential needs are distinguished from desires or wants. Es-
sential needs in this case are illnesses or disabilities defined by medical
criteria and distributed according to a natural lottery. Outka acknowledges
that in many cases the deliberate acts of individuals or institutions may con-
tribute to the development of various diseases; however, he illustrates quite
graphically how complex decision-making might become if we attempt to depart
from the natural lottery model. A second principle that he finds useful in
dec is ion- making is: "Similar treatment for similar cases." However, this
second principle is not germane to the selection of subjects for established
categories of research. Rather, it is employed to determine which categories
of research (or innovative therapy) to establish. He uses it principally to
establish categories of individuals for whom therapy will not be developed.
Thus, Outka agrees with Ramsey (23, at p. 252 et seq) that when one must
decide between claimants for a medical treatment unavailable to all, it might
be most appropriate to develop a policy of random patient selection. He further
agrees that the best suitable alternative might be to deny all persons in certain
4-53
categories specific medical treatment.
I should now like to propose a general approach to developing lottery
systems for recruiting subjects for research characterized as having either
known or unknown risk of physical or psychological harm. This approach has
built into it a device for securing community consultation to determine
whether the group as a group perceives the potential benefits of the re-
search as being sufficient to merit the presentation of the risk to even
a small number of them. Further, entry into the lottery may be validated
only by the informed consent of the individual.
It would be most appropriate to use this form of lottery in the very
early stages of innovative therapy where there is some cause to suspect
that there might be substantial risk. Thus, for example, it might be used
to determine the first persons upon whom an invasive diagnostic technique- -
eg, needle biopsy of the kidney, cardiac catheterization, and so on--is
tried. It would also be appropriate to use this system in research in which
a subject assumes known risk for purposes of developing biological information
about a class of persons with no immediate prospect of direct personal health-
related benefit. Although the primary purpose of developing lottery systems
is to achieve fair distribution of risk, other advantages of this proposal
will be specified.
Procedures
1) A formal description should be prepared of the biological attributes
the prospective subjects must have in order to test the hypothesis.
2) The proposal to conduct the research—including the formal description
of prospective subjects—should be disseminated widely through the community
4-54
upon which it is planned to draw for prospective subjects. This might be
accomplished through advertisement. It is now common practice to advertise
to recruit subjects for basic research. However, recruiting subjects for
innovative therapy may be seen as conflicting with ethical codes of medical
societies proscribing advertisement of physicians' services. Parentheti-
cally, at the time of this writing the Federal Trade Commission is con-
sidering development of regulations that would require physicians to ad-
vertise (24). Alternatively, the research proposal might be circulated
among practicing physicians along with a request that they inform those of
their patients having the needed biological attributes of Che proposal. This
procedure is now commonly utilized by the National Institutes of Health (NIH).
They commonly circulate letters to practicing physicians to recruit subjects
with various disorders. An additional mechanism might involve publication
of announcements of the proposed research in medical journals.
One purpose of such wide dissemination of the proposal is to minimize
the probability of exploitation of the administratively available. Another
purpose is to get the broadest possible base of opinion to serve the ob-
jectives of community consultation.
3) The individuals having the necessary biological attributes should
be informed that there will be a meeting at which the research proposal will
be discussed. They should further be informed that the first purpose of the
meeting is to determine whether the research should be conducted. Further,
they should be informed that they will not be called upon to serve as subjects
without their informed consent; ie, attendance at the meeting does not commit
one to serving as a subject should it be decided to proceed with the research.
4-55
Those who are interested should be requested to contact the individual who
will convene the meeting (ordinarily the investigator).
4) When a sufficient number of prospective subjects have indicated
their interest, the meeting should be convened. At this meeting the in-
vestigator (s) should communicate to the assembled prospective subjects
all relevant details of the proposal. The purposes of this assembly in-
clude those of community consultation. It would most effectively ac-
complish one of the previously specified functions of informed consent;
ie, "To involve the public" (3, at p. 3).
At this meeting the fragility of the individual in relation to the in-
vestigator should be reduced considerably. The prospective subjects will
greatly outnumber the investigators. One might expect that they would
develop a sense of community. They may, for example, decide that the
collective benefit is insufficient to merit the degree of risk to be borne
by even a small number of them. Through the development of this sense of
community, the prospective subject population may develop a sense of co-
hesiveness and mutual support. One of the effects of this might be to
mitigate--to some extent- -.against some of the hazards of depending upon
the volunteer mechanism that have been detailed earlier as some types of
derivative psychological risks (2, at pp. 10-15).
The group should be informed that subjects for the research will be
chosen by a lottery and that no person will participate in the lottery with-
out having first consented as an individual.
4) An attempt should be made to have at least 5-10 times as many
persons volunteer for participation in the lottery than will actually be
4-56
used as subjects. If the group has determined by majority vote that it
is appropriate to proceed with the research yet only a tiny minority of
the group volunteers for participation in the lottery, one might suspect
the validity of the process for that particular proposal. The credibility
of the lottery system would certainly be enhanced if those investigators
having the biologic attributes needed for participation volunteered for the
lottery.
5) In the case of innovative therapy all persons who consent to become
part of the lottery would be promised first access to the innovation should it
prove to be suitably safe and effective. The offer of first access as an "in-
ducement" to participate in therapeutic innovation is not without precedent;
eg, in the initial placebo-controlled trials of polio vaccine, such an offer
was made to the families of the children whose parents consented to their
participation (10, at p. 147).
CATEGORIES OF RESEARCH REEXAMINED
Research presenting mere inconvenience to the subject --2
Guidelines : In research presenting mere inconvenience- -but neither known
or unknown risks of physical or psychological harm- -all humans may be con-
sidered appropriate subjects providing that:
a) The subjects have sufficient capacity to comprehend not only to give
suitable consent to become a subject but also to perform the duties expected of
a subject. Truly uncomprehending subjects should be used only if no other
class of humans possesses the needed biological attributes to test the hypothesis
4-57
of the research; in such cases, proxy consent is required.
b) Refusal to become a subject will not jeopardize a dependent re-
lationship desired by the subject; exceptions may be justified by com-
munity consent of those sharing the dependent relationship. In general,
these exceptions should be allowed only when the proposed research is clearly
in the interests of improving the institution.
There are some features of this guideline that may be perceived by some
as disadvantageous. To the extent that economic inducements are offered to
participate in this type of research, there will tend to be a disproport-
ionately high representation in the subject population of those who are on
welfare and others who perceive themselves as impoverished. To the extent
or implicitly
that psychosocial benefits are offered--explicitly--there will tend to be a
disproportionately high representation of the "administratively available"
and those who question their own self -worth. (I doubt that guidelines can
be developed to resolve the latter problem. However, investigators could
be reminded to avoid inappropriate capitalization on the vulnerabilities of
such persons.)
Research presenting unknown risk of physical or psychological harm - -2
Guideline: In research activities entailing unknown risk all reasonable
effort should be made to conjoin the direct health interests of the subjects
to the ends of the research. This guideline is most easily applicable in
conducting activities that have been defined previously as innovative/
therapy (1, at 9a- 18a).
If this guideline is adopted it would change drastically the ways in
which new drugs are developed in the United States. It would be tantamount
4-58
to saying that most phase I drug testing is inappropriate. It would not
particularly change the approach to research in any other category of in-
novative therapy. For example, we do not now find it necessary to prove
that innovative approaches to surgery, invasive diagnostic techniques (in-
cluding biopsies and catheterizations), diagnostic and therapeutic devices,
use of radioisotopes, radiation and other physical forms of therapy, and
behavioral approaches to psychotherapy are safe in "normal humans" before
we proceed to determine whether they are effective in the populations they
are intended to benefit.
The disadvantages of Phase I testing of new drugs were explored in
detail at a recent conference (25). To summarize some of the arguments:
Azarnoff (25, at p. 796 et seq) concluded: "The wisdom of routine
testing of drugs in healthy volunteers is questioned on two counts: Scien-
tific validity and the dubious ethical practice of giving normal individuals
drugs that are not for their benefit and may be harmful to them. The meta-
bolism of drugs is so changed by numerous diseases that any transfer of re-
sults from healthy volunteers to patients may be dangerously misleading.
Drug interactions also modify results and must be evaluated in both normal
and patient-subjects."
Oates (25, at p. 809 et seq) concluded: "The determination of safety
and adverse effects in Phase I together with investigations on the dis-
position and bioavailability of the drug often provide a better scientific
and ethical basis if conducted with knowledge of the dose required to produce
the desired pharmacologic effect. Because certain pharmacologic effects are
specific for the disease state, Phase I studies on such drugs should be con-
ducted only in carefully selected patients with the appropriate disease."
Hollister (25, at p. 803 et seq), in discussing problems peculiar to
early studies on psychotherapeutic drugs emphasized: "It does not seem
likely that one can predict these (psychotherapeutic) effects in any useful
or innovative way from the study of such drugs in normal, asymptomatic human
subjects."
There are some types of research that present unknown risk which are not
designed to bring direct benefit to the subjects. In general, this type of
research is designed to further our understanding of the biology of humans
4-59
or certain categories of humans (the categories may be determined in a
variety of ways including the possession of a certain disease). At times
it is hoped that based upon improved understandings of the biology of humans
it will be possible to develop an innovative approach to therapy. However,
almost all techniques used to develop improved understandings of biology in
persons with various diseases have known risks and should be treated accordingly.
Let us now consider research procedures having truly unknown risks which
are not designed to bring direct health benefit to particular subjects. In
general, it should be possible to identify prospective subject populations
whose interests most closely approximate those of the research. Alternatively,
it should be possible to identify individuals who experience similar conditions
as will be created for research purposes in the course of their usual acti-
vities. Thus, for example, the first experiments designed to test the effects
of high barometric pressure might most appropriately be done on individuals who
experience such conditions routinely; eg, deep sea divers. Research designed
to determine the physiological effects of prolonged exposure to zero gravity
first became of interest as the plans to develop the astronaut program were
developing. Persons who were to become astronauts would experience zero
gravity conditions for prolonged periods. It would have been impossible to
develop the astronaut program without knowing the effects of zero gravity,
whether it could be tolerated, and what measures could be taken to offset
potentially harmful effects--if any. Thus, it seems that it might be most
appropriate to select as the initial subjects for such studies those who
wished to become astronauts. It might be quite appropriate to inform those
who wish to become astronauts that they will not be accepted unless they are
4-60
willing to assume the unknown risk of research designed to test the phy-
siological effects of zero gravity. This suggests the exclusion stated
earlier to the general formulation for guidelines for the selection of
dependent subjects; viz, there are some rare cases in which research pre-
senting known and/or unknown risk is necessary to establish a category of
employment; in such cases it may be appropriate to offer such employment
contingent upon an individual's willingness to be a subject of such research.
Choices of who shall be the first, and so on, would be resolved most fairly
through a lottery; in this case participation would not be voluntary.
Special cases of unknown risk research : In the course of medical
practice, use of a "standard" prophylactic, diagnostic, or therapeutic moda-
lity may be associated with a severe adverse reaction. When this first occurs,
there will almost always be uncertainty as to whether the modality was the
cause of the reaction. Determinations as to whether the intervention and
the reaction are causally related are necessary to enable one to predict
with reasonable confidence the consequences of repeating that intervention
in similar persons in the future. The traditional approaches to establishing
correlations now most prevelant are as follows: a) The physician may write
a "case report"; physicians who read the case report and who have seen similar
correlations are thus prompted to report their own observations, b) In the
event the adverse reaction is not potentially lethal or disabling, the in-
dividual who first experienced the adverse reaction and others like him may
be exposed deliberately to the putative causative modality to see if the same
adverse reaction recurs, c) Upon first observing an association between an
intervention and an adverse reaction, a careful review may be made of the
4-61
consequences of performing this intervention in similar patients. This may
be accomplished by review of medical records of similar patients.
Illustrative example : At one point it was first observed that the
performance of an intravenous pyelogram (IVP) in a patient with multiple
myeloma was followed by acute tubular necrosis (ATN) . At the time it was
known that ATN (an often severe and sometimes lethal kidney disease) was a
very rare complication of IVP (a diagnostic test usually thought of as re-
latively harmless). Perhaps the first time this was observed, those who ob-
served it did not think to draw a correlation. But at some point sufficient
experience accumulated to permit formulation of the hypothesis as a question:
Are persons with multiple myeloma peculiarly susceptible to this particular
severe complication of this useful diagnostic test? At the point that ac-
cumulated experience indicates that it is legitimate to ask the question,
it becomes necessary to categorize the use of this "standard" diagnostic
procedure in persons with multiple myeloma as innovative therapy with un-
known risk. All reasonable means to test the hypothesis without exposing
new persons to serious risk should be exhausted before proceeding. Thus,
for example, review of the medical records of persons with multiple myeloma
may reveal a high incidence of ATN following IVP. However, for the sake of
developing a guideline, let us suppose that the cause-effect relationship
cannot be established without testing the IVP on additional persons with
multiple myeloma.
Persons with multiple myeloma often develop kidney disease as a com-
plication of their primary disease. However, they may also develop diseases
of the kidney that are not related to their primary disease and which are
4-62
amenable to corrective therapy. The IVP may be most useful in distinguishing
between the two types of disease. Other diagnostic tests may yield infor-
mation similar to that obtainable with IVP. In order to get all of the in-
formation one can obtain from an IVP it is necessary to do several tests
each of which present their own inconvenience and risks. Performance of
some of these tests in individuals who are not peculiarly susceptible to ATN
from IVPs carries with them significantly higher risk than does the IVP. How-
ever, the risk of all of these tests taken together and calculated as a
function of probability plus magnitude of harm does not nearly equal the
risk of say a 10 per cent probability of ATN. Thus, it is in the interest of
the class of persons who have multiple myeloma to learn whether they are ex-
cessively susceptible to ATN following IVPs.
In this situation—and in like situations — it seems most appropriate
to establish a modified form of lottery. In this case, a group of persons
having multiple myeloma might be assembled. They should be informed of the
probability that they might develop some sort of kidney disease and that at
such time it will be in their interest to have appropriate diagnostic testing
done. Then they might have presented to them the known and unknown risks and
benefits of IVP as well as the known risks and benefits of alternative dia-
gnostic techniques. They should be informed that at the time evidence of
kidney disease develops in any of them, one or the other of the alternative
approaches to diagnosis should be used. At this point — without knowing which
among them might develop kidney disease and, of those, which might develop it
first— they should be asked to choose between the alternatives. Some may
choose to "play it relatively safe" assuming the relatively higher known risks
of alternative diagnostic techniques. Those who choose for IVP will be further
4-63
chosen by the natural lottery. Those who choose for IVP should also agree
in advance to have sophisticated tests done for the purpose of detecting
mild or sub-clinical cases of ATN.
In this particular situation, it would be appropriate to disseminate
widely the facts that the question had been raised and the research was being
conducted. This would encourage other physician-investigators to either
establish similar research procedures at their own medical centers or to
refer appropriate patients to centers at which this research was being done.
Even more importantly, it would caution physicians against imposing unknown
risks inadvertantly while awaiting the results of the research.
General formulation : The category of research presenting unknown risk
of physical or psychological harm includes most activities classified as in-
novative therapy (practice) at the very earliest stages of their development.
The most suitable subjects for such research are those who are vulnerable
by virtue of having assumed the sick role. In general, subjects for this
categ ory of research should be the most capable available having the bio-
logical attributes necessary to test the hypothesis. In cases in which there
is some a priori reason to predict serious harm--eg, as in the special case
exemplified by testing IVPs in patients with multiple myeloma- -community con-
sultation should be sought; under some circumstances a modified lottery will
be the most appropriate device to select subjects.
Combinations of known and unknown risk of physical or psychological
harm- -2
Almost all research to be considered in this category is innovative
therapy. However, this is, in general, innovative therapy at much more ad-
4-64
vanced stages than those activities described in the preceding section. Thus,
in general, the high standards for selection of capable subjects called for
in the preceding section will not be required. Devices such as community con-
sultation and lottery systems will almost never be needed. Some special pro-
blems in this category will be exposed through examination of the special case
known as the randomized clinical trial (infra). Earlier, this category was
divided into two large subsets for purposes of developing guidelines.
The first subset is defined by the use of a therapeutic modality proved
effective for a certain disorder in one human population which is now to be
tested for safety and efficacy in different sorts of persons having the same
disorder. It is the custom in the United States to develop new drugs based
upon testing of their safety and efficacy almost exclusively . in adults who are
incapable of becoming pregnant. The reasons for this were presented to the
Commission during its deliberations on research on the fetus and have been
discussed in detail by Capron (20), Mirkin (26), and Shirkey (25, at p. 827 et
seq). As a consequence, most drugs must contain on their FDA-approved package
labels a statement to the effect that their safety and/or efficacy have not
been established in children and/or pregnant women. In fact, it might be
more appropriate to include in such statements that the safety and/or efficacy
of the drug has not been established in women who are capable of becoming preg-
nant.
When a drug that has been proved safe and effective in the adult male and
infertile adult female population is used either in children or in fertile or
pregnant adult females, such use is attended by the risks discovered through
testing in the preceding populations. These are the known risks of such use.
In addition, there may be unknown risks which are discovered only through use
4-65
in- the new populations. A clear case in point is thalidomide which was
found to be very safe and very effective as a sleeping pill in non-pregnant
adults. Other drugs known to be reasonably safe in adults were found to be
highly toxic to infants; eg, chloramphenicol.
It is common practice in this country to administer drugs approved for
use in non-pregnant adults to pregnant women and children. Such administra-
tion is conducted according to the usual standards of medical practice with-
out rigorous testing of safety and/or efficacy. In this way we have a ten-
dency to distribute the unknown risks of such activities not randomly but
rather capriciously. In addition, we have no assurance that, should the
risks materialize as harm, they will be detected.
As noted earlier, in this subset of this category of research, it may
ordinarily be assumed that the benefits identified in other populations of
humans will accrue to the new population in which the drug is used. Thus,
if a pregnant woman or a child is afflicted with an illness that threatens
to cause either death or permanent disability, almost no-one would challenge
the right of that person to be treated with a drug that might produce a cure.
For example, such a person might have a serious infection with an organism
that is sensitive to (capable of being killed or having its growth arrested
by) an antibiotic not proved safe or effective in such persons. It seems
appropriate in such cases to administer the antibiotic.
When the treatment is symptomatic, greater problems are presented to
those responsible for making decisions. If the child or pregnant woman is
not threatened with death or disability, does that person have the right to
assume unknown risk on behalf of himself or her fetus? In other words, does
that person have the moral obligation to endure pain, nausea, insomnia, and
4-66
so on, in order to avoid unknown risk of physical or psychological harm?
Many persons in this category choose to take the symptomatic remedy (26).
Whether they are aware of the unknown risk they are assuming is uncertain;
it seems safe to guess that in most cases they are not.
The dimensions of this problem—particularly as it relates to drug
therapy in pregnant women and children—have been explored in detail by
Mirkin (26). He points out, for example, that the effects of drugs that
influence the central nervous system (eg, tranquillizers, sleeping pills,
sedatives, analgesics, alcohol, and so on) on behavioral development are
virtually totally unknown. Yet, approximately 32 per cent of pregnant women
take tranquillizers during the course of their pregnancies.
Parenthetically, it should be emphasized that such "unapproved" use of
drugs must rarely be associated with substantial harm; if it were commonly
associated we would probably be much more aware of it. What we must be alert
for is the extraordinary toxicity. A mechanism should be devised for early
detection of such toxicities as those associated with administration of
thalidomide to pregnant women and chloramphenicol to infants. Yet, this
mechanism should not deprive individuals in such populations of their usual
expectation of usually equivalent benefit in exchange for usually equivalent
risk.
In general, new drugs--and other modalities where appropriate—should be
tested very carefully in animals in such a way as to be able to anticipate
toxicities peculiar to developing humans. Assuming that animal testing pro-
duces no cause to suspect toxicities peculiar to developing humans, the fol-
lowing general formulation might be appropriate. Although the preceding
discussion has focused primarily on problems peculiar to developing humans,
4-67
the following formulation is intended to be applicable to this entire sub-
set of this category of research.
General formulation : When a therapeutic modality that has been proved
safe and effective for the treatment of a disorder in one class of humans
is to be used for the therapy of the same disorder in another class of
humans, selection of individuals in the new population to receive the
therapeutic modality may closely approximate the standards used in the de-
termination of therapy in the context of practice. The patient- subjects
should be informed that there is, in general, a remote possibility of un-
known physical or psychological harm. They should further be informed
that if they choose to receive the therapeutic modality they incur the
reciprocal obligation of assuming the inconvenience of tests for safety
and/or efficacy in persons like them.
The second subset of this category of research—testing for efficacy
of a therapeutic modality in a disease other than that for which it was
proved safe and ef fective--dif fers from the preceding subset on one im-
portant respect. There are no grounds for equivalent assumptions of benefit.
The standards for selection of subjects for this subset should be closer to
those for research presenting unknown risk than those for the first subset
of this category. Persons having the new disease in which efficacy is to
be tested should be as like as possible the population in which safety has
been established. However, if the new disease for which the modality seems
effective is one which if untreated may cause death or serious disability and
if there are no suitable alternative methods of treatment, it may be ap-
propriate to proceed rapidly to introduce use of this modality in populations
differing substantially from that in which safety has been established.
4-63
The randomized clinical trial (RCT) ; The RCT is a special case of
research which presents combinations of known and unknown risks to the
subjects. The RCT--which is being used with increasing frequency in the
United States--is the subject of an extensive analysis by Fried (10). The
known risks of the RCT are those of the various modalities that are being
compared. The unknown risks include, among other things, the fact that
there is an approximately 50 per cent chance (if two modalities are being
compared) that a subject will learn retrospectively that he has been treated
with the less effective modality. The RCT usually is conducted with the
dual purposes of bringing direct health-related benefit to the patient-sub-
ject and learning how to better benefit like persons in the future.
Fried suggests that, for a variety of reasons, the RCT is used much
more often than is necessary. He suggests that this overuse is related
to the high priority given by scientists to statistical truth—particularly
when they cannot achieve theoretical truth. I shall agree that the device
of the RCT tends to be overused and attempt to develop a guideline that
would tend to limit its use to cases in which it is necessary. If the
assumption that the RCT is overused is not correct, this guideline will
not particularly reduce the frequency with which it is used.
Guideline : An RCT should be conducted only when there exists a legi-
timate question as to which of two or more competing modes of therapy (in-
cluding prophylaxis) is superior. Evaluation of superiority may take into
account either considerations of safety or of efficacy but usually both.
The effect of this guideline would be to rule out the RCT to test the
relative safety or efficacy of a modality against placebo in the treatment
4-69
of conditions for which there exist accepted or standard modes of therapy.
In such cases the new modality should be tested for superiority against
the accepted standard.
Fried suggests that the RCT tends to exploit the administrative avail-
ability of certain types of prospective subjects (10, at p. 63, eg):
"Thus, for instance, Veterans Administration (VA) hospitals have been
particularly apt places for the conduct of RCT's, because of the compre-
hensive nature of the records they keep, the fact that patients moving from
one part of the country to another could be kept within the experiment, and
because administrative coordination between many hospitals is particularly
convenient, thus leading to more valid, general results."
As noted earlier, the RCT has some features in common with social policy
research. This becomes particularly true when an institution such as the VA
offers the RCT as the only mode of therapy for a given condition. Thus, the
following guideline is suggested:
Guideline : Establishment of a RCT within an institution requires that
the prospective subject population have made available to them the alternative
of personal care according to standard and accepted practice within the com-
munity. If the prospective subject population is one which has no reasonable
alternatives available--eg, patients at a Veterans Administration hospital,
military hospital, the only hospital in a town, and so on--and if no alter-
native to participation in the RCT will be offered within the institution,
community consent is required. (However, as noted earlier, in this case
community consent does not override the necessity for individual informed
consent.)
Fried further suggests that in many types of RCT the patient- subject
is obliged to sacrifice whatever he perceives as good in the traditional
physician-patient relationship (10, eg at pp. 160-165). Fried suggests that
4-70
one way of partially resolving this problem is to exclude primary care
physicians from conduct of and participation in RCTs. In his perception,
tertiary care physicians do not ordinarily enter into physician-patient
relationships; consequently, they should be assigned responsibility for
conduct of RCTs including the random allocation of various individuals to
the alternative modes of therapy. I disagree. Many tertiary care physicians
enter into long-term physician-patient relationships. In those sorts of
RCTs in which the services of a tertiary care physician who does not enter
into long-term physician-patient relationships are required, the same de-
gree of short-term involvement with the tertiary care physician would ensure
in the context of either practice or the RCT. Thus, when a surgical pro-
cedure is compared to drug therapy for a condition, the surgeon is likely
to be involved in the care of those referred to him for the same amount of
time whether he is proceeding in the context of practice or RCT.
Guideline: RCTs should be conducted so as to preserve — to the extent
possible—the personal physician-patient relationship. This will ordinarily
mean that the patient- subject can identify one member of the RCT team as
his personal physician to whom he can turn for various aspects of personal
care. Alternatively, the patient-subject should be encouraged to maintain
a doctor-patient relationship with a physician not involved in the RCT but
sufficiently familiar with it to integrate its components and objectives with
those of personal care.
As a corollary to the preceding problem . : Fried suggests that the
system of random allocation to treatment groups may often override the
patient-subject's personal value systems (10, at p. 153 et seq) :
4-71
"Even in medically equivalent cases, patients may have quite different
value systems; their life plans may have quite different structures. And
though the overall prognosis, the overall expected value of the two therapies
may be practically the same, the composition of the risks and benefits of
each therapy might be different. Thus, for instance, surgery for heart disease
in some cases might involve a very high initial risk of surgical mortality
followed by a very good risk for, say, five years of survival after the surgery,
while the standard medical treatment for the same condition might have the same
overall mortality expectation, but with a risk of death distributed much more
evenly over the period of years."
Many other examples might be given and, have been in preceding papers (1-2).
The following guideline is offered as a partial solution to this problem.
Guideline : In the design of a RCT, if there is a reason to believe that
the prospective patient-subject population might have strong preferences between
competing modes of therapy based upon personal value systems and even when the
competing modalities seem medically equivalent, priority should be given to the
personal value systems of the prospective patient-subject.
Thus, if the prospective subject has a strong wish for one of the two com-
peting modalities based upon persona} preference, he should be allowed the
right to select. In the example cited above from Fried, the patient would be
allowed to select between heart surgery and medical therapy. It is feasible--
though less ef ficient--to develop suitable statistical comparisons retros-
pectively based upon matched pairings. Even in such cases a certain number
of prospective patient-subjects may express a wish to be divested of the
burdens of decision-making; accordingly, they may request that their therapy
be decided by lot. When there is any reasonable question that choices between
competing modalities might be based upon personal preferences, community
consultation is advised.
This guideline would present a particular problem to the development of
RCTs which are dependent upon double-blind design. Fried cites one example
4-72
of a RCT that was conducted as a double-blind study in which it seems
to me that some of the patient- subjects might have chosen to be in one
or another group based upon personal value systems. This is the study
of antihypertensive therapy in which one group received placebo and the
other, drugs that were known to produce such side effects as depression,
fatigability, and impairment of sexual function (10, at pp. 144-145). Given
the assumption of equivalent probability of benefit (without which the RCT
is unjustified), it seems to me that many persons would choose not to risk
the known side effects of the drugs. As it turned out, in that particular
study, those who received the drugs not only survived longer but also had
a much lower incidence of some of the serious complications of hypertension.
However, as observed by many commentators, the ethics of a particular re-
search project cannot be determined after the fact based upon the results.
I would still propose that the personal value systems of the individual
prospective subjects should take priority over the requirements of scien-
tific design. Such studies would most appropriately be done on prospective
patient-subjects who either have no choice based upon personal value systems
or wish to be divested of the burdens of such decision-making.
Known risk of physical or psychological harm--2
As this category has been defined it excludes innovative therapy. When
a procedure is performed with the intent of bringing direct benefit to the
individual upon whom it is performed and the risks of the procedure are
sufficiently well known to merit classification by the social device of
accepted or approved (in the population of which the individual is legiti-
mately a member), the procedure is classified as practice. Thus, in this
4-73
category it is necessary to consider only basic research. Ordinarily,
thi3 research will be designed to develop general information about
the biology or behavior of a class of human beings. In some cases, the
results of the research might form the basis .for the development of in-
novative therapy for that class of human beings.
Some guidelines for the selection of subjects for this category of
research were presented earlier (cf, Known risk).
Guidelines : Research presenting known risk of physical or psy-
chological harm to individuals with the expectation that the benefits of
the research will accrue not to the individuals but to the class of humans
they represent should be conducted according to the highest possible
standards of informed consent; only the most capable subjects who will
permit testing of the hypothesis should be selected. If there is any rea-
sonable doubt--as perceived by the IRB--that the class of humans will not
regard the potential benefits as justifying the known risks, community con-
sultation should be required. At the discretion of the community, a lottery
system should be devised for selection of research subjects; no person should
be used as a subject unless he or his legal guardian consents to his parti-
cipation in the lottery.
In some cases community consultation will be desired but the actual
prospective subject population may be so fragile that they cannot be assem-
bled as a group. Alternatively, it may be necessary to initiate the proposed
research on individuals who develop the biologic attributes necessary to test
the hypothesis very shortly before the research must be begun. In such cases,
surrogate consultation may be of value.
4-74
A good illustrative example of a situation in which surrogate con-
sultation might be of value was presented earlier (3, at pp. 48-52). This
proposal presented the known risks of coronary sinus catheterization to in-
dividuals shortly after their admissions to coronary care units with known
or suspected myocardial infarations. The purpose of the research was to
develop information that might be of benefit to the class of persons having
acute myocardial infarctions. Obviously, it is impossible to assemble this
group of people to learn whether they perceive the potential benefits as
meriting the risks to a small number of them.
In this case the surrogate group could be composed of people who had
been discharged recently from coronary care units in which they were hos-
pitalized for treatment of myocardial infarction. Such persons ordinarily
will be aware of the fact that they are at relatively high risk of sustaining
additional myocardial infarctions. This group might be asked if they consider
the potential benefits of the proposed research to be sufficient to merit the
risk presented to even a small number of them. Further, if they were to be
readmitted to a coronary care unit with another myocardial infarction, would
they consent to participation in the research (given the conditions detailed
earlier)?
If a large percentage of the surrogates answered yes to both questions,
the IRB could proceed with confidence to approve this proposal given all of
the precautions cited in the initial description of the project.
SIM-tARY_ AND CONCLUSIONS
Principle : The subjects of research shall have those attributes that will
4-75
permit adequate testing of the hypothesis. In most biomedical and in some
behavioral research the attributes can (and should) be stated precisely in
biological terms. In some behavioral and in most social research, the at-
tributes can (and should) be stated in social terms. An adequate state-
ment of those biological and/or social attributes that establishes a universe
of prospective subjects includes criteria for exclusion as well as inclusion.
This principle establishes criteria for selection of subjects based upon
the requirements of the scientific design of research; however, it has far-
reaching implications in the protection of the rights, health, and welfare
of humans including many who are not subjects of research. This principle
is meant to have priority over any subsequently stated criteria for selection
of subjects.
In many studies it is necessary to include in the subject population in-
dividuals who are commonly called "normal controls" or "healthy volunteers";
there are no such persons. Normality and health are states of being that can-
not be proved scientifically. Thus, it is ordinarily preferred to describe
these individuals as being free of certain specific attributes of non-health
or non-normality.
One exclusion from this principle is suggested by acknowledging that
some types of research have no specific hypotheses (eg, some pilot studies)
and that in some types of research the hypotheses evolve during its early
conduct. Ordinarily, these studies should present no risk ( infra) to the
subjects.
Research categorized by nature of risks
For purposes of developing guidelines, research is categorized by
whether or not it presents risk and, if so, what sort and to whom. It is
4-76
acknowledged that in many cases the nature of both the benefits and the
beneficiaries is of equal importance; in general, the nature of the benefits
is used to establish subsets of research in categories defined by the nature
of the risk. There are five major categories of research determined by risk
criteria:
1) Research presenting mere inconvenience to the subject: Research
which presents the subject with the necessity of bearing the burden of mere
inconvenience is distinguished from that which presents the possibility of
consequential injury (determined as a function of probability and magnitude).
Research in this category employs techniques, modalities, and interventions
that have been tested sufficiently to earn the classification (by the social
device) of accepted or approved (for either research or practice purposes).
Research presenting mere inconvenience is characterized as presenting no
greater risk of consequential injury to the subject than that inherent in
his particular life situation (only as the life situation relates to the
research) .
The next three categories of research present known and/or unknown
risks of physical or psychological harm to individual subjects; in these
categories, possibilities of social, legal, and economic harm are disregarded:
2) Research presenting unknown risk of physical or psychological harm:
Research conducted in this category employs techniques, modalities, or in-
terventions which have not been performed sufficiently often or well in
humans to permit reasonable assessments of the probability and magnitude of
harm they might produce. This category includes most innovative therapies
(practices) at the earliest stages of their development.
4-77
3) Research presenting known risk of physical or psychological harm:
Research considered in this category employs techniques, modalities, and
interventions that have been tested sufficiently well to predict with
reasonable certainty both the probability and magnitude of physical or
psychological harm to the prospective subject population. The probability
and/or magnitude of physical and/or psychological harm is greater than that
inherent in the prospective subject's life situation.
4) Research presenting combinations of known and unknown risk of
physical or psychological harm: The most important type of research in
this category is innovative therapy in which the therapy has been proved
safe and effective in one human population and which is to be tested for
safety and/or efficacy in another. The differences in these populations
are determined by such biological attributes as age, sex, pregnancy, and
altered states of physiology that are either induced deliberately (eg, by
co-administration of another drug) or are naturally occurring (eg, presence
of a different disease).
Three guidelines are proposed for the category of known risk which
are designed to minimize the possibility of physical or psychological harm.
These guidelines are equally applicable to unknown risk research in which
there is some reason to suspect consequential injury; eg, based upon prior
animal testing or toxicities determined in preceding populations.
Guidelines : a) Every possible effort should be made to convert known
risk of physical or psychological harm to mere inconvenience, b) Utmost
effort should be made to identify those individuals most susceptible to
harm for purposes of excluding them from the prospective subject population.
4-78
c) Research should be conducted only by investigators who are highly
skilled in the early detection and minimization of developing or nascent
harms. The facilities immediately available to them to accomplish these
purposes should be the best known.
Before proceeding to discuss either the risks that are borne by groups
(category 5) or the evaluation of benefits from the perspective of the groups
it is necessary to identify four devices that might be utilized.
Community consent is defined as informed consent negotiated with the
community in which it is planned to conduct research or from which it is
planned to draw prospective subjects. The elements of informed consent to
be negotiated with the community are each of those eleven that are germane
to the proposed research.
The community may consent to overriding the necessity for individual
informed consent for several categories of research. These include social
policy research which requires arbitrary allocation of material goods as
well as saturation involvement of the community, and research involving
some types of inconvenience. The community may not consent to imposing
known and/or unknown risks of physical or psychological harm on one of its
members without the consent of that member.
Community consultation is distinguished from community consent; it in-
volves consultation with the community in which it is planned to conduct re-
search or from which it is planned to draw prospective subjects. In general,
the function of community consultation is to assist the investigator in
learning how to better conduct his negotiations for informed consent with
individuals within the community. In cases in which individuals are asked
4-79
to assume known or unknown risk of physical or psychological harm and in
which the potential benefits are expected to accrue more to the class of
persons than to the individual research subject, through community con-
sultation it may be learned whether the class perceives the potential
benefits as being worthy of assuming the risks of research by even a small
proportion of their class. Another purpose of community consultation is to
reduce the fragility of individuals who are either highly vulnerable or highly
dependent in their negotiations with investigators. Community consultation
will also be advised for establishing lottery systems ( infra ) .
Surrogate consent is defined as informed consent negotiated with a group
of individuals that is in all important respects (biologically and/or socially)
virtually identical to the actual proposed research subject population. This
device is used in situations in which approaching the actual proposed subject
population with full information would defeat the purposes of the research.
The elements of informed consent to be negotiated with the surrogates are
each of those eleven that are germane to the proposed research. Surrogate
consent is said to have been achieved when a sufficient number of surrogates
(say, 95 per cent) state that given full information they would have consented
to become research subjects. In contrast to community consent, surrogate
consent can never be used to authorize research in the absence of individual
or proxy informed consent. However, based upon the availability of surrogate
consent, the IRB may decide that it is permissible to proceed with some
categories of research without individual or proxy informed consent; eg, un-
seen observer research.
4-80
Surrogate consultation is distinguished from surrogate consent in
that the surrogates are not virtually identical to the actual proposed
subject population. Surrogate consultation is used in cases in which
community consultation might be desired but the actual proposed subjects
are either too uncomprehending or too vulnerable to provide such consul-
tation. The surrogates should be the closest approximation that is rea-
sonably possible (biologically and/or socially) of the actual proposed
research subject population.
Various procedures may be employed, depending upon the purposes and
and
the populations, to negotiate or achieve community surrogate consent or
consultation; model procedures are proposed in appropriate sections.
5) Risks borne by society, institutions, or classes of people: In
general, when it is a group that will be put at risk, informed consent
should be obtained from the group through community consent.
Social policy research - -conscious and systematic experiments designed
to yield results useful in the formulation of public policy, where this
intervention involves a sample of the human population and sometimes a
control group--often involves the arbitrary allocation of material goods
to a segment of the population and measuring in some formal way the effects
of such allocation on the recipients. Often, a control group that is as
similar as possible is studied to see how well they fare without having
been recipients of the material goods.
Proposed guideline : In social policy research involving the arbitrary
allocation of material goods to one segment of the population and measuring
in some formal way the effects of such allocation on the recipients, community
4-81
consent is required. If there are control subjects (individuals who will
be similarly studied but who will not receive the material goods) they are
to be considered part of the community from which consent is required.
In general, it is assumed that the community (or collection of com-
munities) will be unable to discover any mechanism for allocation of the
goods that they will agree is more just than a lottery.
Some additional guidelines are proposed for selection of subjects for
research that presents risks to groups.
Proposed guideline : In research that presents substantial risks to
society, institutions, or classes of people, community consent is required.
The nature of the research is nearly irrelevant; it may be basic bio-
logical, biomedical, behavioral, or social research. The main criteria for
establishing this class of research is that in the view of the IRB either:
a) The risks to the group in the aggregate are substantial and are greater
than are likely to be perceived by any individual member of the group; or
b) There is likely to be difference of opinion within the group as to
whether the results of the research might be detrimental to the group.
In this category of research there is usually a requirement for negoti-
ating informed consent with each individual. This requirement may be waived
only if both of the following requirements are met: a) There is no conseq-
uential risk of physical or psychological harm to any individual; and b) The
group according to its own established rules of procedure determines that
there is no necessity for each individual to consent.
Proposed guideline for undisclosed purpose research: If in the view of
the IRB, there is reasonable probability that a prospective subject population
4-82
would not consent to undisclosed purpose research if the purpose were
disclosed in advance, the IRB should propose a suitable mechanism for
obtaining surrogate consent.
Proposed guideline for unseen observer research: If in the view of
the IRB, there is reasonable probability that a prospective subject popu-
lation would not consent to unseen observer research if they were made
aware of the fact of its existence, the IRB should propose a suitable
mechanism for obtaining surrogate consent.
Some additional problems are identified in which there might be con-
flicts of interest between individuals or groups as to whether it is approp-
riate to proceed with a particular research proposal; no guidelines are
offered for their resolution. Conflicts of interest may be between society
in general on one hand, and on the other, groups of individuals who might
be perceived as enemies of society. Conflicts are not limited to social
research. For example, in studies of the pathogenesis, prevalence, or
incidence of occupational diseases there are often conflicts between manage-
ment and labor as to whether the research should be done. Another type of
conflict of interest is created when- -in the course of conducting research--
there are unexpected findings that may be so harmful to a group that, per-
haps, they should not be published.
INCAPABLE SUBJECTS are defined as those lacking sufficient power,
prowess, intelligence, or resources to either perform or accomplish: a)
Participation in the research, or b) The protection of their own rights
and welfare. The major focus of the discussion of incapable subjects is
on identifying those who are incapable of protection of their own rights
and welfare. These individuals are further classified as follows: 1) Un-
4-83
comprehending; 2) Vulnerable; and 3) Dependent. These categories are
not exclusive sets; some guidelines developed for the protection of those
in one category are equally applicable for protection of those in another.
The uncomprehending subject is classified as incapable largely by
virtue of being unable to comprehend the information necessary to provide
valid consent. The following are general formulations for the development
of guidelines for the selection of uncomprehending subjects:
a) In those who are intermittently or predictably inebriated or un-
conscious, consent can usually be negotiated in advance. If there is some
cause to suspect that one's judgment might change while inebriated, it
should be negotiated in advance whether the change in judgment is to be
honored. More specifically, the person who expects to become inebriated
may consent to have his judgment while not inebriated take priority over
any expressions to the contrary while inebriated.
b) In developing guidelines to participate in research that presents
mere inconvenience, standards for comprehension should not be set too high.
If they are, it will tend to exclude inappropriately some persons from as-
suming the role of subject as a job. In general, it should be assumed that
a person is capable of suitable comprehension unless this is demonstrably
not the case. When in doubt, comprehension can be assured with the aid of
third party scrutiny or the 2-part consent form.
c) In developing guidelines to participate in research that presents
known and/or unknown risk of physical or psychological harm, standards for
comprehension should be set rather high. To the extent that the benefits of
4-84
such research will accrue more to the class of persons represented by the
subject than directly to the uncomprehending subject there should be an
increasing presumption of the necessity for third party scrutiny or the
2-part consent form.
d) In cases in which the intervention is categorized as innovative
therapy and in which the physician- investigator and the IRB agree that this
is as good as the best alternative therapy (innovative or otherwise) for a
well-defined category of prospective subjects, if neither direct nor proxy
consent is possible, consent may be presumed.
e) The truly uncomprehending subject is, in general, not a suitable
prospective subject for research unless the only individuals that possess
the biological attributes necessary to test the hypothesis are necessarily
also uncomprehending. In such cases it will be necessary to have suitable
proxy consent. In cases in which there might be difficult choices between
reasonable alternatives and in cases in which known and/or unknown risk is
assumed by individuals with a primary purpose of benefitting their class
rather than with an expectation of direct health-related benefit, it may
be appropriate to establish mechanisms for consultation with the community
of proxies. In some cases it might be appropriate to establish mechanisms
for surrogate consultation.
2) Vulnerable subjects are defined as those who are either capable of
being wounded or defenseless against injury. Vulnerable individuals are
categorized further into two groups each deriving its vulnerability from
different origins; these are: A) The sick; and B) The impoverished. A
third category is developed as the potentially vulnerable: C) Minorities.
4-85
A) For present purposes, sickness is defined as a state as perceived
and defined by the individual assuming the sick role. It is a state which
the individual would abandon—with the aid of technically competent help-
in favor of a condition that he would perceive as healthy or at least as
more healthy than he is.
Among those who are vulnerable by virtue of being sick, several cate-
gories are viewed as being highly vulnerable. Persons having prolonged
chronic illnesses which are refractory to standard therapeutic modalities
and persons who question their self-worth are viewed as seriously vulnerable.
Perhaps the most vulnerable are some persons who believe (correctly or in-
correctly) that their own death is imminent.
Recently, it has been proposed that a new class of subject called the
" neomort " might be developed. A neomort is an individual who has been de-
clared dead by virtue of current criteria for establishing brain death and
whose other bodily functions are maintained with the aid of various devices.
Such persons might be maintained indefinitely for a variety of biological
research procedures that for various reasons one might not wish to perform
on sentient beings. Additionally, organs and tissues might be harvested
as necessary for transplant purposes. The neomort proposal is merely
identified as an issue that the Commission might wish to consider; no
guidelines are proposed.
General considerations for the development of guidelines to select
subjects who are vulnerable by virtue of being sick:
a) Persons having prolonged chronic illnesses which are refractory
to standard therapeutic modalities are particularly vulnerable to taking
risks even for remote possibilities of relief when they perceive themselves
4-86
as desperate. Such persons are most likely to be protected by require-
ments that all innovative therapies are to be conducted according to the
highest standards of the relevant scientific discipline. This presumes
that they will first be reviewed by an IRB which will have sufficient
expertise to review the proposed innovative therapy for scientific merit.
b) Persons who question their self-worth may be especially vulnerable
to inappropriate inducement by offering benefits other than those directly
related to their health. Thus, it seems appropriate that the guidelines
reflect a need for caution not to exploit such persons. On the other hand,
advice given to those responsible for selection of subjects should not be
too rigid. Some persons, by virtue of being excluded, might have their
senses of self-esteem diminished further.
c) The most difficult group with which to deal will be those who per-
ceive themselves as dying. Some of these persons are particularly sus-
ceptible to loneliness; inappropriate exclusion of such persons from "normal
human activities" might intensify such feelings. Thus it would be clearly
inappropriate to develop rigid guidelines excluding dying persons from re-
search characterized by mere inconvenience,
In general, research on dying persons which presents known and/or un-
known risk should be limited to that in which the dying person--or at least
that class of persons which he represents—is likely to receive direct
health-related benefits. Experiments designed to use the dying person
merely because there is no further way of causing him damage- -particularly
when these are done without meticulous negotiations for informed consent- -
are an affront to the dignity of such persons.
4-87
On the other hand, guidelines should not be so rigid as to proscribe
altruistic acts done with meticulously negotiated informed consent such
as donation of an organ or tissue according to the provisions of the
Uniform Anatomical Gift Act.
d) Those who are vulnerable by virtue of having assumed the sick role
are, in general, the most suitable subjects for most biomedical research.
To the extent that the research is done with the intent of bringing direct
health related benefits to the subject this seems to present little problem.
A greater problem is presented when sick persons are asked to assume risk
with the expectation that the benefits will redound more to the class of
persons whom they represent than to themselves. In estimating the potential
benefits of research to a class of persons, the investigator is likely to
err, if at all, on the side of overestimating. The individual sick person
tends not to be strategically placed either to disagree with the investigator's
estimate or to refuse an appeal to do good for others. In research presenting
known risk and in some categories of research presenting unknown risk (where
there is some cause to presume that the unknown risk might be substantial)
and the purpose of the research is to develop information that is more likely
to benefit the class than the individual, community consultation is required.
The class of persons who will receive the benefits is better situated than
either the investigator or the individual prospective subject to decide whether
the benefits are sufficiently great that they would wish to expose even a
small number of their class to the known or unknown risks. In some cases it
may be appropriate to establish lottery systems to determine who will be the
subjects of such research ( infra ) . In some cases the proposed subject popu-
4-1
lation may be so fragile that surrogate consultation may be required.
e) In some cases there will be competition for scarce innovative
therapies; the most fair way to determine who shall be the recipients
among those having equal medical needs is a lottery.
B) The impoverished : Impoverishment is defined as a condition in
which a person considers it necessary to take extraordinary risks to secure
money or other economic benefits that will enable him to purchase what he
considers to be the necessities of life. Further, his willingness to take
extraordinary risks is based upon his belief that he cannot secure a suffi-
cient amount of money by ordinary means. This category intersects, but is
not identical with, that of being dependent by virtue of being on welfare.
Proposed guideline : Economic inducements to participate in research
should be determined by the amount of inconvenience to be imposed on the
subject and should not be based on calculations of known or unknown risk.
As a corollary to this guideline it might be stated that research subjects
should not be paid to assume risk; they should rather be compensated for
in j ury .
This guideline, if adopted, would represent a departure from the
tradition in our society that persons are commonly paid high salaries to
assume large risks. Therefore, in order to justify adoption of this guide-
line it will be necessary to provide justification for this departure from
tradition.
General formulation of another guideline to protect the impoverished:
In calculating the costs of health services to an individual who will simul-
4-89
taneously play the roles of subject and patient, it should be calculated
how much his medical care would have cost had he not agreed to play the
role of subject. This amount should be paid by the patient. The ad-
ditional costs incurred as a consequence of agreeing to play the role
of subject should be paid by the investigator or sponsoring agency.
C) Minority groups . Individuals who are members of minority groups
(as determined by race, sex, ethnicity, and so on) are not to be considered
particularly vulnerable a priori. Because the group may be the subject of
discriminatory societal customs, an inordinately high percentage of the
group may be classified as uncomprehending, vulnerable, or dependent by
virtue of other criteria. Such persons should be treated accordingly. The
interests of members of minority groups are probably best protected by re-
presentation on the IRB.
3) The dependent subject is defined as one who is unable to exist,
sustain himself, or act suitably or normally without the assistance or
direction of another or others. He is connected in a subordinate relation-
ship to another person or institution.
In considering the problems of developing guidelines for the selection
of dependent subjects, discussion is focused on dependent relationships that
present either of two potentials. The first is that by virtue of the re-
lationship, the dependent individual is administratively more available to
the investigator to be selected as a subject than are other individuals not
having the same dependent status. The second is that in which the dependent
individual might fear that he might forfeit either in part or in whole his
dependent status by virtue of refusing to become a subject. Related to the
second potential- -but also having its roots in the barriers to autonomy
4-90
identified in connection with the sick role--is the concept of the fragility
of the individual who is either highly dependent or highly vulnerable (by
virtue of having assumed the sick role) . The notion of fragility suggests
the relative lack of power such persons have in negotiating with investi-
gators.
General considerations for the development of guidelines : In general,
it is appropriate to draw upon the administratively available to assume in-
convenience. However, in research presenting risk of physical or psycholo-
gical harm, as the degree of risk increases, there should be a decreasing
assumption of the validity of drawing upon the administratively available.
In situations in which known risk is presented in order to develop basic
of
information that will be presumed benefit to the class, there will be a de-
A
creasing assumption of the appropriateness of drawing upon the administratively
available without consulting them as a group. Thus, in some cases it will be
necessary to seek community consultation and in others, community consent.
In general, it is not appropriate to threaten individuals with loss
of dependent status owing to unwillingness to assume known and/or unknown
risk of physical or psychological harm; rare exceptions are specified sub-
sequently (infra).
In general, it should be assumed that most persons can choose to receive
an innovative therapy. However, the choice to receive an innovative therapy
imposes upon the individual the reciprocal obligation of assuming the in-
convenience of tests necessary to prove its safety and/or efficacy. In cases
in which the innovative therapy is scarce, refusal to assume this reciprocal
obligation may justify exclusion of that individual. However, a distinction
4-91
must be made between depriving a person of a single modality on the one
hand and, on the other, of an entire dependent relationship. It is not
appropriate to terminate a professional-client relationship because the
client refuses to be the subject of research—including innovative therapy.
The issue of developing a system for compensation of harmed research
subjects is reviewed in some detail. Development of a suitable system
would yield several salutary effects to the entire research process. It
will be most difficult to develop appropriate guidelines for the selection
of subjects for research presenting known and/or unknown risk without having
first developed suitable systems for compensation. The proposals for guide-
lines for selection of subjects contained in this paper are based on the
assumption that appropriate mechanisms for compensation of harmed subjects
will be developed.
A proposal for the development of lottery systems to distribute some
sorts of known and unknown research risks is presented. The proposal has
the following features. Built into it is a device for securing community
consultation to determine whether the group as a group perceives the
potential benefits of the research as being sufficient to merit the presen-
tation of the risks to even a small number of them. It is suggested that
an assembly of prospective subjects would tend to reduce the fragility of
the individual negotiating as an individual with the more powerful pro-
fessional. This system might also be used to minimize the possibility of
exploitation of the administratively available. Entry into the lottery
would be validated only by the informed consent of the individual.
4-92
It would be most appropriate to use this form of lottery in the
very early stages of an innovative therapy in those cases in which
there is some cause to suspect that there might be substantial risk.
It would also be appropriate to use this system in research presenting
known risk with no immediate prospect of direct health-related benefit.
Categories of research reexamined
Proposed guideline : In research presenting mere inc onven ience --but
neither known nor unknown risk of physical or psychological harm--all humans
may be considered appropriate subjects providing that:
a) The subjects have sufficient capacity to comprehend not only to
give suitable consent to become a subject but also to perform the duties
expected of a subject. Truly uncomprehending subjects should be used only
if no other class of humans possesses the needed biological attributes; in
such cases, proxy consent is required.
b) Refusal to become a subject will not jeopardize a dependent re-
lationship desired by the subject; exceptions may be justified by community
consent of those sharing the dependent relationship. In general, these ex-
ceptions should be allowed only when the proposed research is clearly in
the interests of improving the institution.
Most research presenting social or legal risks to the individual will
not become manifest as harm unless there are violations of confidentiality.
In most cases, safeguards of confidentiality can be established to the extent
that all concerned can be very confident that there will be no violations.
4-93
Thus most research dealing with confidential information can be considered
as presenting mere inconvenience. Where there are barriers to preservation
of the confidentiality-legal or otherwise—the research might be considered
as presenting known risk.
General formulation for guidelines: In research presenting social or
legal risks to the individual which might become manifest through violations
of confidentiality it is the responsibility of the investigator to inform the
prospective subject of the extent to which confidentiality can be assured.
If the investigator can assure absolute protection of confidentiality the re-
search is to be considered as presenting mere inconvenience. In such cases
the investigator assumes responsibility (liability) for any social or legal
harm done to the subject as a consequence of failure to maintain confiden-
tiality.
If there are barriers to preservation of confidentiality these should
be disclosed to the prospective subject. The prospective subject assumes
the risk of the consequences of violations of confidentiality through
mechanisms of which he has been forewarned.
Some potential disadvantages of these guidelines are discussed.
Some additional guidelines for research presenting unknown risk of
physical or psychological harm are presented; In research activities en-
tailing unknown risk all reasonable effort should be made to conjoin the
direct health interests of the subjects to the ends of the research. This
guideline is most easily applicable in conducting activities that have been
defined as innovative therapy.
If this guideline is adopted it would change drastically the ways in
which new drugs are developed in the United States. It would be tantamount
4-94
to saying that most Phase I drug testing is inappropriate. It would not
particularly change the approach to research in any other category of in-
notative therapy.
Some sorts of research present truly unknown risks and are not de-
signed to bring direct health benefit to particular subjects. In general,
it should be possible to identify prospective subject populations whose
interests most closely approximate those of the research. Alternatively,
it should be possible to identify individuals who experience similar con-
ditions as will be created for research purposes in the course of their usual
activities. An example is given of a type of research that might be con-
ducted in this category that is necessary to establish a new class of employ-
ment. This suggests the exclusion to the earlier stated general formulation
for guidelines for the selection of dependent subjects. There are some rare
cases in which research presenting known and/or unknown risk is necessary to
establish a new class of employment; in such cases it may be appropriate to
offer such employment contingent upon an individual's willingness to be a
subject of such research. Choices of who shall be the first, and so on,
would be resolved most fairly through a lottery; in this case, participation
in the lottery would not be voluntary.
Some special cases of unknown risk research are presented: These de-
velop when, in the course of medical practice, use of a standard prophylactic,
diagnostic, or therapeutic modality seems to be associated with a severe ad-
verse reaction in a class of persons. Further, it is impossible to determine
whether this class of persons is peculiarly susceptible to the severe com-
plication without further research in this group. This special case is ac-
commodated in the following general formulation.
4-95
General formulation for guidelines for research presenting unknown
risk: The category of research presenting unknown risk of physical or
psychological harm includes most activities classified as innovative therapy
(practice) at the very earliest stages of their development. The most suitable
subjects for such research are those who are vulnerable by virtue of being
sick. In general, subjects for this category of research should be the most
capable available having the biological attributes necessary to test the
hypothesis. In cases in which there is some a priori reason to predict
serious harm--eg, as in the special case cited above- -community consultation
should be sought; under some circumstances a modified lottery system will be
the most appropriate device to select subjects.
Research presenting combinations of known and unknown risk of physical
or psychological harm: Almost all research considered in this category is
innovative therapy. However, this is, in general, innovative therapy at
much more advanced stages than those activities characterized as presenting
unknown risk. Thus, in general, the high standards for selection of capable
subjects called for in that category should not be required. Devices such
as community consultation and lottery systems will almostnever be needed.
One subset of research in this category is defined by the use of a
therapeutic modality proved effective for a certain disorder in one human
population which is now to be tested for safety and efficacy in different
sorts of persons having the same disorder. In consideration of selection
of subjects for this subset it is argued that decisions may be based upon
assumptions of equivalent benefit accruing to the new population. Also, in
4-96
this subset, there is a usual assumption of nearly equivalent risk. While
there is no way to predict the extraordinary unknown harm, such harms are
rare. On these bases the following general formulation is proposed.
General formulation: When a therapeutic modality that has been proved
safe and effective for the treatment of a disorder in one class of humans
is to be used for the therapy of the same disorder in another class of
humans, selection of individuals in the new population to receive the
therapeutic modality may closely approximate the standards used for the
determination of therapy in the context of practice. The patient-subjects
should be informed that there is, in general, a remote possibility of un-
known physical or psychological harm. They should further be informed that
if they choose to receive the therapeutic modality, they incur the reciprocal
obligation of assuming the inconvenience of tests for safety in persons like
them.
The second subset of this category of research- -testing for efficacy of
a therapeutic modality in a disease other than that for which it was proved
safe and effective--dif fers from the preceding subset in one important respect.
There are no grounds for equivalent assumptions of benefit. The standards
for selection of subjects for this subset should be closer to those for re-
search presenting unknown risk that those for the first subset of this category.
Those persons having the new disease in which efficacy is to be tested should
be as like as possible the population in which safety has been established.
The randomized clinical trial (RCT) is a special case of research
presenting combinations of known and unknown risks to the subject. Careful
examination of this special case provides the basis for some proposed guide-
4-97
lines some of which may be applicable to other categories of research pre-
senting known and/or unknown risk of physical or psychological harm.
Proposed guidelines : a) An RCT should be conducted only when there
exists a legitimate question as to which of two or more competing modes of
therapy (including prophylaxis) is superior. Evaluation of superiority may
take into account either considerations of safety or of efficacy but usually
both.
The effect of this guideline would be to rule out the RCT to test the
relative safety or efficacy of a modality against placebo in the treatment
of conditions for which there exist accepted standard modes of therapy.
b) Establishment of a RCT within an institution requires that the
prospective subject population have made available to them the alternative
of personal care according to standard and accepted practice within the com-
munity. If the prospective subject population is one which has not reasonable
alternatives available--eg, patients at a Veterans Administration hospital,
military hospital, the only hospital in town, and so on--and if no alter-
native to participation in the RCT will be offered within the institution,
community consent is required. (However, as noted earlier, in this case com-
munity consent does not override the necessity for individual informed consent.
c) RCTs should be conducted so as to preserve--to the extent possible--
the personal physician-patient relationship. This will ordinarily mean that
the patient- subject can identify one member of the RCT team as his personal
physician to whom he can turn for various aspects of personal care. Alter-
natively, the patient-subject should be encouraged to maintain a doctor-patient
relationship with a physician not involved in the RCT but sufficiently familiar
4-98
with it to integrate its components and objectives with those of personal
care.
d) In the design of a RCT, if there is a reason to believe that the
prospective patient- subject population might have strong preferences between
competing modes of therapy based upon personal value systems and even when
the competing modalities seem medically equivalent, priority should be given
to the personal value systems of the prospective patient-subject.
Mechanisms for the implementation of these guidelines, their impli-
cations, and some of the problems they create are discussed.
Research presenting known risk of physical or psychological harm--as
it has been defined in this paper—excludes innovative therapy.
Proposed guideline : Research presenting known risk of physical or psy-
chological harm to individuals with the expectation that the benefits of the
research will accrue not to the individual but to the class of humans they
represent should be conducted according to the highest possible standards of
informed consent; only the most capable subjects who will permit testing of
the hypothesis should be selected. If there is any reasonable doubt--as
perceived by the IRB--that the class of humans will not regard the potential
benefits as justifying the known risks, community consultation should be re-
quired. At the discretion of the community, a lottery system should be de-
vised for selection of research subjects; no person should be used as a
subject unless he or his legal guardian consents to participation in the
lottery.
In some cases, community consultation will be desired but the actual
prospective subject population may be so fragile that they cannot be as-
4-99
sembled as a group. In other cases, it is necessary to initiate the re-
search on individuals who become very ill or otherwise develop the bio-
logical attributes necessary to test the hypothesis very shortly before
the research must be begun (thus, there is no time for community consul-
tation). In such cases, surrogate consultation may be of value.
rt s\ r\ rt Vt ft s\
In developing guidelines for the selection of individual research
subjects, the most difficult problems are presented by those types of re-
search that present substantial risk to the individual with the hope of
developing information that will be of benefit primarily (or only) to the
class of humans he represents. Such research presents two very difficult
questions. The first--which we now have.no way to answer--is:. Does the
class of humans consider the hoped for benefits sufficient to merit the
risk to be borne by even a small fraction of its members? Assuming that
this can be answered affirmatively, the second is: Which individuals shall
assume the risks?
The devices and procedures proposed in this paper for obtaining the
opinions of groups are designed to answer both questions. Ordinarily, if
the answer to the first question is yes, the second would be resolved most
fairly through a lottery.
It must be understood that these devices and procedures are untested.
They may or may not accomplish the purposes for which they are intended.
Even if they do, they might be so cumbersome and expensive that they might
not be worth using. If it is decided to implement these suggestions, their
initial trials should be conducted in such a way as to provide an accurate
4-100
assessment of their costs and benefits.
Meanwhile, these devices and procedures — if they are to be used at
all--should not be used capriciously. They should be used only to re-
solve legitimate and important problems. The necessity for their use
should be determined by the IRB on a protocol by protocol basis.
Finally, the vast majority of research should call upon reasonably
capable subjects either to assume only inconvenience or, alternatively,
to accept reasonably predictable risks in exchange for reasonably pre-
dictable direct health benefits. Thus, in most cases there should be
no need to solicit group opinion (or consent) or to establish lottery
systems to select research subjects.
4-101
References
1. Levine, R.J. : The boundaries between biomedical or behavioral re-
search and the accepted and routine practice of medicine. July 14, 1975;
Addendum: September, 24, 1975.
2. Levine, R. J. : The role of the assessment of risk-benefit criteria in
the determination of the appropriateness of research involving human sub-
jects. October 27, 1975.
3. Levine, R. J. : The nature and definition of informed consent in various
research settings. December 1, 1975.
4. Feinstein, A.R. : Clinical Judgment , Williams and Wilkins Co., Baltimore,
1967, 414 pp.
5. Pauker, S.G. : Coronary surgery: When, where and for whom? N. Eng . J.
Med . 293: 1369-1370, 1975.
6. Report of the Conference on the Use of Stimulant Drugs in the Treatment
of Behaviorally Disturbed Young School Children. DHEW, Washington, D.C.,
January 11-12, 1971.
7. Katz, J. and Capron, A.M. : Catastrophic Diseases : Who Decides What ?
Russell Sage Foundation, New York, 1975, 273 pp.
8. Beecher, H.K. : Research and the Individual : Human Studies. Little
Brown and Co., Boston, 1970, 358 pp.
9. Veatch, R.M. : Ethical principles in medical experimentation, in
Ethical and Legal Issues of Social Experimentation , ed. by A.M. Rivlin
and P.M. Timpane, The Brookings Institution, Washington, 1975, pp. 21-59.
10. Fried, C. : Medical Experimentation : Personal Integrity and Social
Policy . American Elsevier Publishing Co., Inc., New York, 1974, 177 pp.
11. Rawls, J.: A Theory of Justice , The Belknap Press of Harvard Uni-
versity Press, Cambridge, 1971, 607 pp.
12. Levine, R.J. : Ethical considerations in the publication of the
results of research involving human subjects. Clinical Research, 21:
763-767, 1973.
13. DeBakey, L. : Ethically questionable data: Publish or reject? Clinical
Research , 22: 113-121, 1974.
14. Makarushka, J.L. : The requirement for informed consent in research on
human subjects: The problem of uncontrolled consequences of health-related
research. Clinical Research, 24: 64-67, 1976.
4-102
15. Webster's Third New International Dictionary, Merriam Co., Spring-
field, Mass., 1971.
16. Gaylin, W. : Harvesting the dead, Harper's Magazine , September,
1974, pp. 23-30.
17. Katz, J. : Experimentation with Human Beings , Russell Sage Foun-
dation, New York, 1972, 1159 pp.
18. Spiro, H.M. : Constraint and consent--0n being a patient and a sub-
ject, N. En£. J. Med. 293: 1134-1135, 1975.
19. Adams, B.R. and Shea-Stonum, M. : Toward a Theory of control of
medical experimentation with human subjects: The role of compensation.
Case Western Reserve Law Review , 25: 604-648, 1975.
20. Capron, A.M. : Legal considerations affecting clinical pharmacological
studies in children. Clinical Research , 21: 141-150, 1973.
21. A report of the Artificial Heart Assessment Panel of the National
Heart and Lung Institute: The Totally Implantable Artificial Heart DHEW
Publication No. (NIH) 74-191, June 1973, 250 pp.
22. Outka, G. : Social justice and equal access to health care. Per-
spectives in Biol , and Med. 18: 185-203, 1975.
23. Ramsey, P. : The Patient as a Person , Yale Univ. Press, New Haven,
1970, 283 pp.
24. Drug Research Reports 18 (No. 52): 4, December 24, 1975.
25. Proceedings of the first Deer Lodge Conference on Clinical Pharma-
cology, Conference Chairman, L. Lasagna: Clin . Pharmacol . Expt' 1 . Therap .
13': 769-840, 1972.
26. Mirkin, B.L. : Drug Therapy and the developing human: Who cares?
Clinical Research 23: 106-113, 1975.
4-103
II
BASIC ETHICAL PRINCIPLES RELATING TO
RESEARCH INVOLVING HUMAN SUBJECTS
ETHICAL PRINCIPLES AND THEIR VALIDITY
Kurt Baier, D. Phil .
ETHICAL PRINCIPLES AND THEIR VALIDITY
(1) The Problem. The National Commission for the Protection of Human
Subjects was asked to "identify the basic ethical principles which should
underlie the conduct of biomedical and behavioral research involving human
subjects." The task assigned to me is (i) to provide guidelines for identi-
fying ethical principles and ethical reasoning generally and for distinguish-
ing them from other kinds of reasoning which also bears on the acceptability
or inadmissibility of human actions; (ii) to provide guidelines for identify-
ing such other kinds of principles and reasonings, e.g., those employing legal
or technical, esthetic or political considerations, and to explain the inter-
relationship between all these types of principles and reasonings; and (iii)
to describe and comment briefly on the range of thought among contemporary
philosophers regarding the identification of principles and the evaluation
of their validity.
The task presents formidable difficulties of selection and organization,
particularly since it is to be accomplished within the confines of a paper
of reasonable length and readability. I have made every effort in this brief
survey and critique of the great variety of approaches and mutually conflict-
ing, confusing, meagre and inconclusive contemporary writings on the pro-
blems raised by the Commission, not to give undue weight and credibility to
those views with which I myself sympathize. All the same, given the current
state of the field, my own perspective on the problem inevitably will not
only have colored my own answers, however qualified and non-committal, to
these difficult questions, but probably also even my selection and presenta-
tion of the materials examined.
5-1
The essay is organized under four heads. Section I contains a brief
survey and critical discussion of the way a number of contemporary philo-
sophers, in what is generally regarded as the mainstream of Anglo-American
thought on the subject, have viewed the problems posed. From this discus-
sion it emerges that moral reasoning is now widely viewed as a special form
of practical reasoning, but that there are major disagreements about pre-
cisely how this form of reasoning is to be characterized, and even about
whether any kind of objective or neutral investigation can settle these dis-
agreements. The remaining three Sections, II-IV, are given over to the
task of providing as objective or neutral a characterization as is possible
of moral reasoning and of how it differs from other forms of practical rea-
soning. More specifically, Section II is a preliminary but indispensable
excursus into the nature of practical reasoning; Section III sets out the
peculiarities of moral reasoning with occasional asides, where apposite,
about its significant differences from neighboring forms of practical rea-
soning; and finally, Section IV, attempts to provide whatever objective
guidelines it seemed possible to give about how to recognize ethical prin-
ciples and test their validity, as well as how to avoid confusing ethical
with other types of practical principles.
I. Current Views
(2) There is general agreement that certain forms of words, "Non-
therapeutic experimentation on aborted fetuses is morally wrong," "To experi-
ment on these children putting them at risk for the sake of future gains to
humanity was unjust," "The neglect of one's professional duties is repre-
5-2
hensible," "it was wrong of you to criticize me in front of these people"
normally express moral judgments. But there is a good deal of disagree-
ment and doubt about others, e.g., "knowledge is intrinsically good,"
"honesty is the best policy," "you are impertinent," "that was a mean thing
to do."
(2.1) Ethical Terms and Ethical Judgments. It is therefore natural
to ask what it is in such judgments that makes them ethical. During this
century four major answers have been given to this question. Some, e.g.,
2 3
G.E. Moore, or W.D. Ross, locate it in the fact that these judgments attri-
bute to things certain quite extraordinary properties, with which ethics is
properly concerned. The main objections to such theories are that it is
impossible to detect or ascertain in any reliable way the presence of these
unusual properties and that even if that were not so, the mere fact that
things or even actions have certain properties does not of itself provide a
reason why anyone should therefore act in one way rather than another. But
to provide such reason is surely the main business of ethics. And so this
account of what it is that makes a judgment ethical is, at best, inadequate.
Therefore, some later theorists, such as the Emotivists and Prescripti-
vists, rejected peculiar ethical properties and located the ethical in the
peculiar function of ethical judgments: the Emotivists in their expressive
and evocative, the Prescriptivists in their imperatival or action-guiding
function. The two main objections to these theories are (i) that they mis-
leadingly restrict the ethical to one or other of the many linguistic func-
tions all of which are performed by ethical judgments in one context or another,
5-3
but which are equally performed by non-ethical judgments. (Thus, while "You
thief!" typically expresses feeling, it does not guide action, and while "You
ought to pay not later than four weeks after delivery" guides action, it
does not express feeling nor evoke it, and "If he had not delivered on time,
you need not have payed him" does neither.) (i) They leave out what is an
essential feature of the language of ethics, namely, that it comprises a form
of practical reasoning, leading to conclusions which may but need not provide
action-guidance for some particular agent. (Thus, "If she had become preg-
nant, it would have been his moral duty to marry her" hints at some under-
lying moral reasoning but comes too late to provide guidance.)
A fourth school, therefore, locates the ethical in the peculiar form
of reasoning on which certain judgments are based. One of the most widely
acclaimed theories, that of R.M. Hare, combines elements of the third and
fourth schools. He maintains that what makes judgments ethical is the fact
that they are prescriptive and universal izable, and that these two features
generate a peculiar form of reasoning, requiring its own peculiar logic, the
logic of imperatives.
It seems fair to say that there is, at this time, very widespread agree-
ment that it is at least a necessary, if not a sufficient condition, of a
judgment's being ethical, that it should be capable of being supported by a
certain method of reasoning. The main disagreements within this broad consen-
ts
sus concern the precise nature of this method of reasoning.
(2.2) Theoretical and Practical Reasons. It is generally agreed that
moral reasoning is practical as opposed to theoretical, and that the difference
5-4
g
is of considerable importance. It is, for instance, often claimed that
G.E. Moore's ethical theories are vitiated by his failure to realize that
the business of ethics was practical rather than theoretical, that ethical
judgments had practical functions, such as action-guiding, and that to the
extent to which they were capable of being supported by reasoning, the rea-
soning had to be practical rather than theoretical. The difference can be
put in this way. Practical reasons are practical in two important senses.
In the first place, they are practical in the sense of being "justificatory,"
or as I prefer to say, "normative," rather than "explanatory" or "probative."
I prefer normative to justificatory because practical reasons are used not
only in justification but also in deliberation. Thus, the reaons we employ
in practical reasoning are reasons to (i.e., reasons why someone should ) be-
lieve, desire, want, feel, prefer or do something, rather than (explanatory)
reasons why , someone in fact believes, desires, etc., something, or why
something has happened, has certain properties, or exists, or (probative)
reasons showing that these things are as claimed. They are, therefore, reasons
concerning matters over which we have some direct or indirect control, and
employed in contexts in which there is a question of how to exercise this con-
trol, rather than a question of simply taking cognizance of the matter.
These reasons are, however, practical also in the sense that they are used
in determining what someone should cto rather than what he should believe . The
reasons employed in such reasoning are reasons for action , that is, they are
only one particular sub-class of the larger class of normative reasons.
The assumption that moral thinking, reasoning, and arguing is practical
in this double sense rests on the further assumption that the central and most
5-5
important question in ethics is practical, i.e., (what someone should do,
morally speaking, or) what, morally speaking, is the right thing for him to
do. This assumption is made by the classical and dominant tradition in
ethical theory and I shall accept it here without argument. But there are
9a
thinkers who regard as central another question, namely, what is a morally
good person? On such an approach, it is not so clear that moral reasoning
and thinking is essentially and primarily practical in the sense explained.
(2.3) Meta-ethical inquiries into the nature of morality. There is,
however, a second line of inquiry which intersects with and complicates the
line already sketched. It sets out from the thought that, though the, moral
thinking of individuals and groups has peculiarities distinguishing one such
morality from another, yet such social and individual moralities are recog-
nizable as moralities , and distinguishable from legal systems, religions,
customs, manners, and so forth. Why not then, this line of inquiry insists,
look for the common and peculiar characteristics of such recognizable morali-
ties, if we want to find out what, if anything, is the function of moral
judgments, moral reasoning, and moral talk generally?
We can divide the responses to this question into three groups. The first
is skeptical. It says that "morality" has so many different senses that no
useful purpose can be served in unraveling them.
The second maintains that "morality" is, not univocal, though the number of
senses is manageably small, so that there is something to be gained by clari-
fying them. Most frequently only two senses are found, a social and a per-
sonal one. Morality in the first sense is a system of generally known and
5-6
applicable principles, rules, or precepts, usually tied to a particular social
order, which those subject to it recognize as binding on them. In the second
sense, a morality is the set of principles, precepts, or particular dictates
coming from one's own conscience, reason, intuition, or moral sense, which one
recognizes on that account as absolutely binding on one, though not on anyone
else.
The third concedes that "morality" may have several senses--! t admits,
for instance, that while in one sense it is the opposite of "immorality," in
a second sense it contrasts with "lay," "religion," "manners," etc., --but
maintains that the relevant (i.e., the second) sense is univocal, and that
with proper care, an investigation of that sense can bring to light the true
nature and function of morality. However, those who take this last line then
disagree with one another on what such a careful, unbiased investigation does
bring to light. We can distinguish three major positions.
12
The first yields the widest sense of "morality." On this view, a morality
is that set of principles, rules, precepts and dictates which the society or
individual in fact recognizes as supreme and as overriding all others. It has
been objected to this view that it lets in too much. On this view, it is said,
law and religion and a person's strongest commitments (e.g., to obey the
Fuhrer, to destroy the Jews), all count as part of morality, and this is im-
plausible.
The second view' 3 says that in listing the essential characteristics of
a morality we are defining "morality." But if such a definition is to be
"neutral," and not simply a covert and illegitimate way of smuggling in sub-
5-7
stantive moral conclusions on the pretence of linguistic investigation, then
the definition has to be purely formal. That is to say, it must not include
as part of the definition any object or goal or purpose or function of
morality. For that would be to abandon the neutral stance appropriate to one
who is trying to illuminate the domain of morality, when this is not the
opposite of immorality, but of law, religion, art, etiquette, expediency, or
wisdom. As an instance of this type of theory we can again mention that of
R.M. Hare. For he holds that we can say of a person, or a group, that he or
it has a morality if and only if they have a set of principles, rules, or
precepts which they regard as universal izable and prescriptive, as well as
overriding.
The main objections to the two additional criteria, have been these.
The objections and their sympathizers maintain that there are moral judg-
ments, such as those of supererogation, which are not universal izable. If
I conclude that I ought to give 50% of my income to charity I do not neces-
sarily imply that everyone else in my position ought to. I may realize that
we cannot demand these things of each other but only of ourselves. The other
objection has come from those, who like the Ethical Egoists, maintain that
universalizability is not a truly formal, because not a truly neutral criterion,
For some if not all versions of Ethical Egoism preclude the universalizability
of ethical judgments. If as an Egoist I argue that I should make an exception
in my favor, I am not implying that everyone else should so do also, i.e., make
an exception in his favor for that might, and normally would, not be in my best
interest. Those who live by Ethical Egoism are therefore excluded from having
a morality. Their Egoistic reasoning is, by this definition, implied to be
5-8
overridden by conclusions of universalizable, i.e., ethical reasoning. In
this way a substantive ethical conclusion is derived from this definition
of "morality," which is not therefore a truly formal one. Any definition
of "morality" which adds any criterion to that of overridingness is thus
non-neutral in that it gives a higher place in the hierarchy of types of
practical reasoning to the type that satisfies these additional criteria
than to the types that do not.
The main objection to including prescriptivity in the definition of
"morality" is as follows. If prescriptivity means no more than guiding ac-
tion, then it is innocuous but also superfluous. For a morality must in
any case contain directives which people can use in solving their practical
problems. If, however, it means more, e.g., what Hare means by it, then it
is objectionable because it mischaracterizes moral judgment. Hare means two
things by "prescriptivity." The first is that if someone makes a prescrip-
tive judgment, then he tells someone to do something, he intends the person
to whom it is addressed to do what he tells him to do. But this is simply
not true of moral judgments. Thus, if on hearing the full story, I now
judge in my mind that a certain person morally ought to have told the full
truth in a certain matter, then this certainly does not imply that I now
intend that other person to have told the truth (for that does not even make
sense), nor does it imply that I intended him to do so at the time when, as
I now judge, he ought to have told the truth.
15
The second thing Hare means by "prescriptivity" is that if someone
accepts a prescriptive judgment addressed to him, then he must act in accord-
5-9
ance with it or at least set himself to do so. And this also is false. For
I may quite honestly think that I ought to do something here and now, say,
terminate life support to a terminal patient and yet not do it, nor intend
to do it, nor set myself to do it.
However, the most serious objection to formal definitions such as Hare's
1 fi
comes from a third type which offers "material" definitions. Such theories
maintain that "morality" must be defined as in some sense social. Some de-
fine "morality" in terms of a social function, such as the promotion of social
harmony or the promotion of general welfare. Others say that to be a morality,
a set of principles and rules must contain at least directives for the settle-
ment of interpersonal conflicts of interests, and directives requiring every-
one's consideration of the interests of at least some others than himself.
On such material definitions of "morality" certain sets of principles
would not be considered moralities, even if they satisfied the three formal
criteria mentioned in the previous discussion. Consider the following three
principles: (i) Always to get out of bed left foot first, (ii) Never to
pick one's nose in public, (iii) To surround oneself as much as possible with
beautiful objects. Now if someone treated these principles as supreme, uni-
versalizable, and prescriptive, then this set would satisfy the three criteria
of Hare's formal definition but, supporters of material definitions would
argue this would not be a morality. To many people, including myself, this
seems a telling point.
The most interesting and deepest-cutting difficulties, however, have
come to light when one examines various material definitions. Such examina-
5-10
tions put the spotlight on another belief about morality which has been com-
monplace at least since Kant. It is the belief that moral judgments are
rational not only in the sense that they are valid conclusions of certain
forms of argumentation, but also in the sense that they provide anyone with
adequate reason to act in accordance with them and against acting contrary
to them, even if they run counter to our inclinations and our interest. This
is one of the implications of the Kantian claim that they are "categorical
imperatives." But this belief causes problems for material definitions.
For it seems that if one accepts a material definition, then the question
of whether one has reason to follow the conclusions of one's moral arguments
becomes a contingent matter. For one then does not necessarily have reason
to do what such judgments require, particularly when they conflict with one's
self-interest or one's desires. If the function of morality is the promotion
of social harmony, and if I do not care much about social harmony, then I may
often have no adequate reason to follow the principles of morality rather
than those of self-interest or simply my desires.
One way out of this difficulty, which has recently been suggested, is to
drop the categorical ity requirement in definitions of a morality. This still
leaves us free to regard moral principles and judgments as not merely spelling
out ways of attaining our preexisting ends. We can think of them as somewhat
like the rules of chess which we have reason to follow because we want to
play chess, or the canons of good manners, because we want to be well mannered.
In these cases, we follow these rules not for the prizes which following them
may gain us, but for its own sake, the sake of doing that doing which consists
in following these rules. On this model, the rules of morality are the con-
5-11
stitutive rules of an "autotelic" activity.
However, this way out is open to a serious objection. It turns morality
into a practice with principles and precepts which a person has reason to
follow only if he wants or likes to engage in that sort of activity. If he
does not, there is no reason why he should follow these principles. Yet,
when we think about the principles and judgments of a morality, it does seem
plain to us that we have reason to demand of others and they of us that we
all adhere to these principles.
A second way out is to say that the principles of morality and the canons
of self-interest always coincide. But apart from making morality unnecessary,
this solution is quite certainly false in the world as we know it. Our brief
examination of Prisoner's Dilemma in Sections II and III below, should make
this plausible. It would be a third way out of the difficulty if it were the
case that the basic ethical principles are such that from their very nature
they ought to be regarded as overriding the reasons of self-interest. If
this way out were possible, we could think of the principles and precepts
of morality both as categorical and as satisfying all the requirements of
formal theories. Whether they could also satisfy the material requirements
would depend on precisely what content, if any, we can or must give to moral
principles which can satisfy the requirements of categoricality.
If the line of thought sketched so far is sound, then it is most plausi-
ble to conceive of a morality as a form of practical reasoning transmitted
and taught by societies, for use by individuals in the solution of their
practical problems, beginning (in deliberation) from general principles to
5-12
specific conclusions solving personal practical problems, conclusions which
are rightly regarded as overriding the conclusions arrived at by all other
forms of practical reasoning, including self-interested reasoning.
Out next aim should therefore be to provide a characterization of the
relevant features of practical reasoning in general and the place of self-
interested reasoning in it. This would put us in a position to see whether
there really are any principles which from their very nature ought to be re-
garded as overriding all other reasons including reasons of self-interest,
and if so, whether there are any other reasons for accepting the account of
moral reasoning implied by this third line of thought.
II. Practical Reasoning
(3.1) Thinking without language. There is a kind of practical thinking
which does not depend on the thinker's having a language or having grown up
within a culture. Many of the higher animals are generally believed to be
capable of this kind of thinking or problem-solving. They are assumed to be
capable of perceiving certain situations in which they find themselves, as
posing "practical problems" and certain developments in which they are involved,
as "solutions to these problems." A cat in a puzzle box trying to find its
way out, a squirrel eyeing a bird feeder, Koehler's famous ape reaching for
the bananas outside his cage, would be examples of animals in such recognized
problem situations. One can say that what makes these situations problematic
for these animals is that they want to get hold of or do something but that
they do not know how . Their practical thinking consists or manifests itself
5-13
in their intelligent goal -directed behavior, whether their intelligence shows
itself merely in the recognition of a development as a solution and in the
ability to learn from their experience to improve the problem-solving per-
formance (as in the case of cats in puzzle boxes or rats in mazes), or whether
intelligence is also exhibited in the economy or even ingenuity of the attempted
solutions (as in the case of Koehler's ape).
Such practical thinking is severely limited in scope. What the indivi-
dual has learned cannot be communicated to others. There is therefore no
scope for advice, or for joint and thus more effective thinking. Each indi-
vidual and each generation must start from scratch. They cannot profit from
the experience of others, or only in the most limited way, e.g., demonstrating
how something is done, as when a cat shows her kittens how to trap and catch
a bird.
(3.2) Practical Reasoning. The presence of a language introduces into
practical thinking four powerful new factors for growth. The first is the
possibility of accumulating and storing practical knowledge based on the ex-
perience of many different individuals and even generations in a given group
culture.
The second factor is making this knowledge available to group members
through education in the form of publicly available general directives, such
as principles, rules, maxims, precepts, and the like, and by teaching them
the skill of applying them to their individual practical problems.
The third factor, arising directly out of the second, is the availability
of practical knowledge with the status of intersubjective validity. An indivi-
5-14
dual learning to apply the publicly available general directives to himself
and his special situation, learns pari passu to apply them, mutatis mutandis ,
also to others in relevantly similar situations. He must learn to move in
thought from the most general principles, such as "Always do unto others as
you would have them do unto you," "Always do what is in your own best interest,"
"Never give a sucker an even break," and so on, down to "therefore, do this
here and now," whether it is his own case or that of another. He must, in
in other words, learn what are the relevant similarities and differences between
the different cases. Obviously, if there is to be such practical thinking,
the general directives on which it rests will have to be formulated in such a
way that they are capable of being employed by individuals of different types
and in different situations. Such general usability of a general directive
will be a condition of incorporation in the culture of a group. By "practical
reasoning," we ordinarily mean this sophisticated practical thinking based
on such general directives. It is because animals lack such general directives
that the practical thinking of animals does not amount to practical reasoning.
The fourth new factor is the possibility of a conflict between reason,
i.e., the outcome of such practical thinking and inclincation. A person en-
gaging in practical reasoning, may arrive at conclusions, i.e., answers to the
question of what he should do, which direct him to do something he is not
inclined, or is inclined not, to do. Thus, I may find, in the mousetrap I
have set up in the kitchen, a mouse injured but still alive. My strong incli-
nation may be to run and leave the mouse to its fate. However, practical
reasoning based on certain principles to which I subscribe, e.g., "Never
inflict or allow avoidable suffering," may lead me to conclude that I ought
5-15
to take the mouse out of the trap and kill it quickly and painlessly. Again,
my inclination may be to light a cigarette, but reasoning based on principles
such as "Never unnecessarily endanger your health" may lead me to conclude
that I ought never to smoke again.
(3.3) Theoretical and Practical Tasks. Concerning each practical ques-
tion, "What should N do?", there is thus what we might call a "theoretical"
and a "practical" task. The theoretical task consists in applying the rele-
vant publicly available general directives to the problem in hand, and ter-
minates in a particular practical judgment of the form "N should do A," where
this has a degree of specificity sufficient to enable N to translate the judg-
ment into action. "N ought never to smoke again" or "N ought to take the
mouse out of the trap," etc. "N ought always to do others as he would have
them do to him" does not have for normal adults the requisite specificity.
There is an obvious but important difference between N's theoretical
and his practical task. The former can be performed by anyone with the rele-
vant knowledge, whether he is N himself or someone else. But the practical
task can be performed only by N himself. If, let us say, N's wife takes the
mouse out of the trap for him and kills it quickly with a well-aimed blow of
a hammer, she either has not performed N's practical task or the practical task
was not as described. If N ought to have taken the mouse out, her taking it out
is not the performance of N's practical task. If her taking it out was the
performance of the practical task in hand, then the outcome of the theoretical
task must have been, not "N should take out the mouse etc.," but perhaps "N
should see to it that the mouse is quickly taken out (by whomever)" or " Some -
one in the kitchen should take out the mouse"etc.
5-16
Once there are publicly available general directives for use in practi-
cal thinking then there is the possibility that N arrives by a correct appli-
cation of these principles to his case, at the judgment that he ought to do
A, but then does something else instead. Thus N may think that the publicly
recognized general directives applicable to his case are principles of sound
practical reasoning and he may in applying them to his case arrive at the
same judgment that anyone else, applying them correctly would arrive at,
namely, that N ought to do A, yet he may then not do what he recognizes as
having the weight of reason behind it.
(3.4) Considerations and Reasons. We must therefore distinguish between
what I shall call "practical considerations" and "practical reasons," respec-
tively. Suppose I come to believe both F-i , that smoking causes lung cancer,
and ?2> that giving up smoking will disrupt my work. And suppose also that
when I attend to F-i , I am inclined to quit smoking, but when I attend to Fo,
I am inclined to continue. Then F-j is for me a consideration in favor of
quitting, and F2 a consideration in favor of continuing. But now suppose that
when I attend for some time to both F-j and F~, then I decide to give up smok-
ing. Then although F] and F2 are for me considerations relevant to whether or
not to give up smoking, F-| is a weightier consideration than F2, for when
attending to both, F-| inclines me more strongly in favor of quitting than F2
inclines me in favor of continuing.
Reasons differ in an important respect from considerations. If F-j is a
reason to quite smoking, then if I know or believe F-j , I ought to be inclined
to quit smoking, quite irrespective of whether I actually attend to it ade-
quately, and whether, supposing I do so attend, I then am so inclined. And
5-17
similarly, if F-j is a weightier reason for giving up smoking than F2 is for
continuing, then I ought to be more strongly inclined to quit than to conti-
nue; and I ought to be so inclined quite irrespective of whether I adequately
attend to both F-| and F2, and irrespective of whether, supposing I so attend,
I then am so inclined.
There is thus no contradiction in saying that F2 is for N a weightier
consideration than F, , but F, is a weightier reason than Fp. It is even possi-
ble that F-j is for N a consideration against, although it is in fact a reason
for, quitting. What makes a fact a consideration and a weightier one than
another fact, is thus a totally different matter from what makes a fact a
reason, and a weightier reason than another fact. Making a mistake about what
is for someone a consideration in favor of doing A, is simply a mistake about
his psychology: about what would weigh with him, and how strongly, in what
circumstances. Making a mistake about what is a reason to do A is not such a
simple psychological matter, but presupposes a different sort of theory about
what makes facts reasons. To believe, correctly or mistakenly, that a certain
fact constitutes a reason to do A, implies that such a fact ought to weigh
with one in favor of doing A, that one ought therefore to be disposed to do A,
even if one does not come naturally to be inclined to do A simply by attending
to this fact. Obviously, such a theory of reason is designed to give practi-
cal reasoners guidance about what to do which is based on a knowledge of what
is a reason, and not on some psychological self-experiment. We can now turn
to a discussion of a few popular theories of practical reasons.
(4) Theories of Practical Reasons. Theories of this type are tied to
standards of rationality. We judge a person or his behavior to be rational
5-18
or irrational according as he, or it, does or does not measure up to certain
demands about the performance of the theoretical and the practical tasks.
Before indicating what these standards are, we must mention some features
of this practice of evaluating people by these standards.
(4.1) Performance and social pressure. People have an interest in know-
ing about any given individual with whom they have dealings, how well he per-
forms his theoretical and practical tasks. For just what dealings they may
be willing to enter into with him may depend on just how reliable is his per-
formance of these tasks. It is, now, therefore surprising that we find a
widespread practice of evaluating others on this basis and, when asked (some-
times even when not asked), to transmit such judgments to third parties. In
view of the recognized importance of this task, people also monitor their
own performance and try to maintain the required standards of rationality.
In some types of practical reasoning, morality being the most obvious
case, society will not be satisfied with letting the individual choose whether
to keep his performance up to the mark and reap the benefits or let his
standards slip and take the consequences. In these cases society attempts
in various ways to ensure the general maintenance of the minimal standard.
The two most widely used methods are training and sanctions. We not merely
teach the young the difference between right and wrong but also try to in-
culcate in them "a good will," that is, a readiness to perform their tasks
adequately, even in the face of contrary inclination. The imposition of
sanctions has brought in its train the widespread practice of requiring prac-
tical reasoners to justify themselves, that is, to be able to rebut accusa-
tions that they have not adequately performed their tasks. We shall note that
5-19
though in all forms of practical reasoning, there will be a tendency, on
account of the importance of the matter, for practical reasoners to justify
themselves when they are accused--no one likes to be thought a fool or a
knave--it is only in relation to certain forms of practical reasoning, morality
being again the prime case, that success at justification is a condition of
not suffering the relevant sanctions. We do think that in moral matters it
is not solely up to the individual reasoner whether or not he wants to per-
form the theoretical and practical tasks adequately. By contrast, though we
have an interest in knowing how imprudent or foolish our fellows are, we
would hardly wish to insist, or allow others to insist, that members of the
community be forced to be less imprudent or fool ish--except perhaps in those
special cases in which important interests of others are also affected.
(4.2) Perfect Rationality. We can now return to the various standards
of rationality, by which we assess the performance of practical reasoners.
The sense of "rational" which is relevant to this investigation is the oppo-
site of "irrational," not "non-rational."' 8 In this sense, a rational, i.e.,
a not irrational, being is one whose reasoning does not fall below a certain
standard of excellence. Ordinarily this standard is quite low. One has to
perform very poorly indeed to qualify as irrational. Thus, failing to make
the effort to rid oneself of one's tendency to overeat when one is overweight
and has a heart condition which is adversely affected by overweight, is con -
trary to reason but not necessarily irrational . The reasons against making
the effort--say, it would disrupt one's work--may be quite weighty enough
to save one's failure from being irrational, even though of course these
reasons will not be as weighty as those in favor of making the effort.
5-20
There is, however, another sense of "rationality," namely, "perfect
rationality" which implies a higher standard. In this sense it is used of a
person whose behavior is always in accordance with reason, behavior which
the agent has adequate reason to think has the weight of reason behind it.
In this paper, "rationality" always means perfect rationality .
However, "rationality" in this sense cannot be applied to anything un-
til the notion of a practical reason has been further clarified. Up to
now, we have skirted the central problem, namely, that of spelling out what
it is to believe that a fact about a certain course of action constitutes
a reason to enter on that course of action. I briefly sketch three widely
held theories of practical reason:
(4.3) Informed Inclination. The simplest kind of theory, and one
which appears to be widely held, 19 at least by implication, simply declares
that what are, for a person, considerations are also reasons for him. That
is to say, rationality consists in always acting in accordance with one's
informed inclination. It should be noted that this is a genuine theory of
reasons, not merely a rejection of reasons in favor of considerations. The
difference is important, for belief in considerations does not commit one
to anything beyond psychological explanatory theory. A theory of reasons
by contrast, is an evaluative theory (Cf. above (2.2)). It ties the ex-
cellence of a person's performance of the theoretical and practical tasks
to the evaluation of a person as rational or irrational, or in other ways
excellent or defective.
5-21
On this simple view, rationality need never involve acting contrary to
inclination. Such unpleasantness can always be avoided, at least in princi-
ple, by taking the trouble to attend adequately to all the facts. Now, of
course, this may be a different if not an impossible task. And if it cannot
be performed, then an individual who knows what facts are for him considera-
tions, might still have to act contrary to his current inclination, since if
he were able adequately to attend to the facts, his inclinations would pre-
sumably be different from what they now are.
What kind of assistance could a culture on this view give to practical
reasoners? It could provide general practical directives of how people of
certain types ought to act. The culture would simply spell out what are for
what (psychological) types of persons, considerations in favor and against
what in what sorts of circumstances, but would present these considerations
as reasons.
This simple theory has several weaknesses. For one thing it suffers
from the practical difficulty or impossibility of saying clearly what it
is for a person to attend adequately to a given fact. For how long and in
what manner must a person attend? But in the absence of such a criterion,
we cannot tell what are for a given person considerations.
More interesting is the fact that such reasons are "self-anchored," in
a sense still to be explained, and that the theory is therefore exposed to
some of the objections to Rational Egoism, namely, those connected with
Prisoners' Dilemma, which will be examined shortly (below, (4.4)).
5-22
(4.4) Rational Choice. A second theory of rationality is that found
in Decision Theory.' 99 There rationality is conceived as an evaluation of
the relationship between a person's choices and his preferences. The question
asked is, " Given your preferences , is it rational for you to choose A?" The
principles of rationality are therefore those of rational choice. The most
frequently cited principles of rational choice are (i) The principle of effec-
tive means: given that one has a certain object, one is to choose the best,
i.e., the most economical means, (ii) The principle of inclusiveness: given
that one has several aims or desires, one is to choose that plan which is most
inclusive, i.e., promises to attain the greatest number of these objects, (iii)
The principle of greater likelihood: given that one has several objects whose
realization by the available means is not equally likely, one is to choose the
plan which is the most likely to succeed.
This notion of rationality leaves no room for the question of whether the
person's preferences, aims or objectives are themselves rational. If rationality
is entirely a matter of the ways in which we can achieve ends, then the choice
of ends or the possession of preferences themselves is not a matter of reason.
Since it is quite natural to go on to ask whether such aims or preferences are
morally justifiable, it is then only logical to exclude morality from the realm
of rationality. It should be obvious, however, that this exclusion follows
only because of this rather limited (though highly influential) conception of
?l
practical rationality.
(4.5) Rational Egoism. A third theory offers a single criterion of
what is a reason for doing something. On this view, a supposed fact about doing
5-23
A is a reason for N to do A if and only if doing A unfavorably affects N's
interest, and indifferent (not a reason'one way or the other) if and only
if doing A does not affect N's interest either favorably or unfavorably.
If F-j is a weightier reason than F2, if and only if when N is confronted by
a choice between doing A and doing B, doing A would more favorably affect
his interest than doing B.
As a theory, Rational Egoism is in some respects more satisfactory than
the two preceding ones. Even though it is difficult to say exactly what is
meant by "favorably affecting someone's interest," we understand the idea
well enough to employ it successfully in many cases. And it does not seem
impossible to refine the concept and build up a body of practical knowledge
vastly more detailed and useful than is our present conventional widsom on
the subject. Lastly, the theory is complete, in the sense that there are
no practical questions or problems about which we could not in principle
find an answer along the lines of Rational Egoism.
The theory does, however, have a glaring weakness, to which I have
already referred. As has frequently been pointed out if people interacting
with one another subscribe to this theory and know or believe of one another
that they so subscribe, they thereby deprive themselves of certain benefits
and so bring it about that their interest is more adversely or less favorably
affected than it would be by other policies. One type of situation in which
this occurs, "Prisoner's Dilemma," has been widely studied, and can be briefly
described.
;-24
Two prisoners are accused of a joint bank robbery, as well as minor crimes
separately committed. The prosecutor tells each that if they both confess
the bank robbery, both get 10 years; if the first confesses and the second
does not, the first gets off free, the second gets 20 years; and if neither
confesses, they both get only 1 year (for their minor crimes). Supposing that
their preferences are strictly parallel to the length of time in prison, then
the preferences (or interest) of each ranked from highest to lowest are: (i)
himself confessing and the other not confessing, i.e., himself getting off
free, the other getting 20 years, (ii) neither confessing, thus both getting
1 year; (iii) both confessing, thus both getting 10 years; (iv) himself not
confessing, and the other confessing, thus himself getting 20 years, the
other getting off free.
Clearly, on the theory of Rational Egoism, it is rational for each to
confess, for whatever the other does , he will do better if he confesses than
if he does not. If the other confesses, then by confessing he will get 10
rather than 20 years. If the other does not confess, then by confessing he
will get off free rather than getting off free should the other be foolish
enough not to confess.
Since it is rational for both to confess, both will get 10 years. They
are therefore both doing less well than if both refused to confess. But,
being Rational Egoists, this policy is not available to them.
What can the Rational Egoist do in such a situation? He can do two things,
(i) He can, following Hobbes, work out certain policies analogous to the
policy of nonconfession which, if but only if universally adopted , would benefit
5-25
everybody more than if the policies of Rational Egoism were universally
adopted. Let us call such policies "distributively beneficial," as op-
posed to those of the Rational Egoist which are "reflexively beneficial."
The difference is that a person benefits from others following distribu-
tively beneficial policies, not from himself following them. The Rational
Egoist will therefore not follow such policies. To derive the benefit
of distributively beneficial policies, Rational Egoists must, therefore,
(ii) create a social order, or support an already existing one, in which
distributively beneficial rules are imposed as overriding the members' own
individual policies of Rational Egoism, and supported by effective sanctions ,
In this way, the distributively beneficial rules are also made reflexively
beneficial, and so can be followed by the Rational Egoist.
Thus, by creating a compulsory social order in which distributively
beneficial rules are enforced, the Rational Egoist can derive these benefits
not available to him in what Hobbes called the state of nature. Of course,
he will have to pay his share of the cost of the enforcement machinery. To
the extent that a particular Rational Egoist can detect a loophole in the
social order by which he can escape the social sanction, he may even reap
the double benefit (at someone else's expense) of breaking the social rule
and of doing what is reflexively beneficial. He is here in the position of
the prisoner who knows that the other will not confess and who can confess
with impunity himself, thus getting off free, though at the expense of the
other who will get 20 years.
The consequence is that the Rational Egoist has reason to set up and
maintain a social order containing a system of distributively beneficial
5-26
and enforceable rules, but he cannot regard these rules as themselves con -
stituting reasons for him to act. He has reason to obey these rules only to
the extent to which obeying them coincides with following egoistic reasons.
Where, through a failure of the sanctions, egoistic reasons diverse from
what the rules require, it would be contrary to reason for him to follow
rather than to break the rules. A theory such as the one I have just sketched
23
may have been held by Bentham.
This is an impressively simple, coherent, and plausible theory of reasons.
In particular, let it be noted, it does not introduce practical reasons of
more than one kind, and so has no problem about the relation between egoistic
and moral reasons. Nevertheless, it has two quite serious weaknesses. The
first is its tendency towards absolutism, a point clearly seen and perhaps
welcomed by Hobbes. For a community of Rational Egoists who know or suspect
each other of Rational Egoism will have (egoistic) reason both to break the
social rules when it is to their advantage to do so, and to ensure that others
will never be in a position to do so. They will therefore want to ensure by
the most stringent policies of "law and order" that crime does not pay other
people . At the same time they will attempt by bribes and threats and other
methods to bend the law to their own advantage. And the officials whose task
it is to ensure that the rules are enforced will if they are Rational Egoists
try to enrich themselves by bending the law, for a consideration. Of course,
there nowhere are or have been communities of Rational Egoists, though it is
perhaps not difficult to envisage a community in which the proportion of
Rational Egoists is steadily growing so that those who are not must assume of
more and more of their fellows that they are.
5-27
The second weakness is that in such a community of Rational Egoists,
there is no conception of justice: no rational limit to self-aggrandise-
ment, as Plato pointed out in the Republic. Every member of this community
will therefore want to change the laws and other compulsory rules in a
direction which will promote his advantage even if it is at the expense
of others. And he will do it in e\/ery way, by bribes, threats, and the
like, which promises success and does not unduly expose him to the risk of
having the sanctions imposed on him. Thus, while the members of a community
remain Rational Egoists, and cannot therefore recognize the compulsory rules
of the community as constituting reasons to act in the way these rules re-
quire, they will continue to have towards each other the same egoistic atti-
tudes they had in the state of nature. They will regard the rules of the
social order as no more than obstacles in the race for self-aggrandisement.
They must therefore attempt to enlarge these obstacles in the path of others
and reduce them in their own paths.
It seems clear that even Rational Egoism, though an impressive theory
of reasons, is by no means wholly satisfactory. The three theories so far
sketched do, however, make clear that the task of a theory of reason is to
yield principles following which will promote the good or, even optimal,
life for people within communities, in which there are publicly recognized
distributively beneficial rules which are, at least most of the time, treated
as overriding the reflexively beneficial canons- of self-interest. It is also
plain that the worst shortcomings of Rational Egoism could be remedied by
a theory of reason from which it followed that these distributively beneficial
rules themselves constituted practical reasons to abide by them and sufficiently
5-28
weighty ones at that to override the canons of self-interest when the two
come into conflict with each other.
Our brief survey of the activity of practical reasoning and of what is
involved in saying that a certain supposed fact is a reason for someone to
do a certain thing, has brought to light two important facts. The first is
that practical reasoning is, from its nature, designed to override what
someone is inclined to do, even when he is fully informed of the facts that
affect his inclination. The second is that any account of practical reason-
ing is committed to some theory of what a reason is, i.e., what makes a fact
a reason for someone to do something. Not even such intuitively plausible
claims as that the fact that doing something is in one's best interest is
a reason for him to do A is more than a hypothesis, and as our discussion of
Prisoners' Dilemma showed, it is not one that is without difficulty. The
main difficulty, of course, arises out of the fact easily overlooked that a
person's interest is affected not only by what he does himself, but also by
what other people do and by developments (such as a flood or the exhaustion
of certain resources) which are not anyone's doing or at any rate not the
planned outcome of human action. Therefore, if one wants one's interest af-
fected as favorably as possible one has to take into account more than one's
own policies. Hence Rational Egoism, which does not allow coordinated poli-
cies, deprives itself of one of the most effective means of getting one's
interest promoted without promoting it oneself. For, as we have seen, by
following distributive^ beneficial rules one does not in any way promote
one's own interest, but if such rules are generally followed, one's interest
is thereby promoted.
5-29
This flaw in the theory of Rational Egoism can be remedied if we do not
construe our theory of reasons, the way Rational Egoism does, as the strong
claim, that the fact that a course of action is in one's interest and only
that fact is a reason for one to enter on that course. None of the diffi-
culties sketched will arise if we drop the second half of this claim and say
instead that although the fact emphasized by Rational Egoism really is a rea-
son, it is not the only fact that is. If, for instance, we were to add to
our theory the claim that the compulsory social rules, which are needed to
combat the worst consequences of Rational Egoism and other self-anchored sys-
tems of reasoning, are also reasons, namely, reasons to act as these rules
require, and reasons which override the reasons of self-interest, then even the
secondary difficulties resulting from the general acceptance of Rational
Egoism would be avoided. But, of course, such a modification would give rise
to many other difficulties. For one thing, while it is plausible to say
that, at least other things equal, it is rational to do what is in one's
best interest, there is no plausibility whatever, especially for someone
who finds Rational Egoism plausible, in this modification of Rational Egoism.
In the following section, on moral reasoning, we shall examine what a
morality is, what the peculiarities of moral reasoning are, how it differs
from self-interested reasoning, and whether the peculiarities of moral rea-
soning are designed to overcome the difficulties we have noted with self-
interested reasoning, and whether it succeeds in doing so.
III. Moral Reasoning
(5) The Parts of a Culture. Our two main tasks, the clarification of
5-30
moral and other kinds of practical reasoning will be advanced if we begin
by looking at the various domains of a culture in which these other kinds of
reasoning are characteristically employed.
We can usefully divide the culture of a society into its theoretical,
its practical, and its aesthetic part. Nothing will here be said about the
aesthetic part. In the theoretical part belong its entire accumulated ex-
planatory knowledge and belief, primarily its pure sciences. In its prac-
tical part belong its applied theoretical or technical knowledge or know-how,
that is, its applied sciences, crafts, and technologies; its practical wis-
dom, that is, its canons of the good life; and, finally, its social order or
mores, that is, the directives spelling out what the society requires or
expects of its members. We shall need to ask what part or parts of a culture
its morality belongs in.
(6) Technical Reasoning. This type of reasoning starts out from a set
of possible ends, such as health, wealth, shelter, nutrition, clothing, power,
transporation, and so on. Such possible ends are assumed, on good evidence,
to be the actual ends of many people on many occasions in many different cir-
cumstances. In technical reasoning, it is merely assumed, not established
that it is always or indeed ever rational to try (or not to try) to attain
any of these ends. Technical reasoning is, therefore, a truncated or limited
form of practical reasoning, a mere stretch or patch of it. The main idea
of technical knowledge is to make publicly available such reliable stretches
of practical reasoning, leaving it to individual practical reasoners in their
individually different situations to determine for themselves whether it is
>-31
rational for them to try to attain such an end.
Suppose a patient suffers from severe mental deterioration and memory
deficit. Suppose his condition is diagnosed as occult hydrocephalus, which
causes decreased mental abilities by interference with absorption of cere-
brospinal fluid. The relevant part of the technical reasoning of the surgeon
could be formulated as follows:
(?) My task (end) is to cure this patient,
(ii) This patient suffers from mental deterioration,
(iii) The cause of his mental deterioration is the interference with
the absorption of the cerebrospinal fluid which is brought on
by occult hydrocephalus,
(iv) If I am to cure the mental deterioration, I must stop the inter-
ference with the absorption of the fluid,
(v) If I am to stop this interference, I must place a plastic tube
through his skull to drain the cerebrospinal fluid from the
brain to the vascular system.
(vi) Therefore I must place a tube . . . etc.
In this reasoning, it is of course taken for granted that the task or end,
curing a patient, is a rational one. Steps (ii) and (iii), diagnosis and deter-
mination of causal factors, are the conclusions of medical thinking, perhaps
reasoning. This part belongs in the scientific, theoretical part of a culture.
Steps (iv) and (v) also rely on theoretical knowledge. However, these
theoretical ingredients can be removed. Then these steps have the form, "If
I am to . . . , I am to . . , ." Depending on what exactly is the theoretical
background knowledge available, "I am to ... " can be correctly replaced by
5-32
"I could ... ," or "I must ... " or "I ought to ... ." If the only known way
to stop the interference, etc., is to insert a tube, then "I must" is the
correct replacement. If there are various ways, "I could ... " is the correct
one. And if inserting a tube is the best of several ways, then "I ought ... "
is the correct one. In this last case, where there are available alternatives
the principles of rational choice (Cf. above (4.4)) determine which is the
best.
Technical reasoning thus has two important features. It starts from pre-
mises which are either general, i.e., possible ends, or are particular in-
stances of such possible ends, a particular person's (Jones') end. These pre-
mises state ends whose rationality in general or in a particular case is no
more than assumed, or implied. And such technical reasoning ends with direc-
tives which are sufficiently specific to enable a person to translate them
into action.
(7) Politics. A particularly important case of technical reasoning
occurs in the political domain. Politics is possible only in a society
which has a political order, that is, one with political institutions, that
is institutions providing for the roles of government and governed. The
most important general functions of government are the regulation of what
people must and need not do for one another and what they must not and may
do to one another, the organization and coordination of cooperative enter-
prises for socially important purposes, such as defense, education, trans-
portation, welfare, and the regulation of privately initiated cooperative
enterprises. Politics is the activity of trying to influence the actions
and policies of government.
5-33
One type of reasoning in this domain starts from particular political
ends, that is, ends of bringing about particular governmental policy changes,
or maintaining particular governmental officials in power. Applied political
science tries to provide the premises and principles needed for this type
of reasoning.
However, what normally goes by the name of political argument, is typi-
cally a moral argument about what governmental policy ought to be. The prin-
ciples needed for this kind of argument are usually discussed in political
philosophy, not in political science, though the dividing line has again be-
come somewhat fluid.
(8) Practical Wisdom. Practical reasoning aimed at practical wisdom
differs from technical reasoning in that it is complete and unlimited. The
conclusions at which it arrives are neither hypothetical nor truncated. They
are not hypothetical, as would be an argument beginning with "Suppose my end
is ... ." And they are not truncated, as those of technical reasoning are,
for they try to establish that it is rational to do a certain thing, which
means that, if in doing it one has a certain end, having that end is also
rational. Typically, such reasoning has a form something like the following.
(i) It is irrational to be imprudent.
(ii) It is imprudent to do something for which one may have to pay
damages.
(iii) One has to pay damages if one is convicted in a malpractice
suit.
(iv) One may be convicted in a malpractice suit if one conducts
5-34
experiments on a fetus scheduled for abortion, and the
mother changes her mind about aborting.
(v) Therefore it is irrational to conduct such experiments.
The main difference between technical reasoning and this form of reason-
ing is that the aim of the former is to identify manageable ways of attaining
possible or actual ends without looking into the rationality of these ends,
the aim of the latter is primarily to arrive at judgments about the rationality
of a given enterprise or project within the entire life plan or system of
inter-related life plans. Technical reasoning goes over short stretches of
frequently trodden ground, yielding detailed, reliable and economical in-
structions of how to get to these destinations. Practical reasoning which
aims at practical wisdom yields conclusions about whether it is wise to
choose certain destinations, given one's overall situation and life plan, and
those of one's fellows.
(9) Mores, Custom, and Law. The third sector of the practical part of
a culture is its mores or social order. In the sense in which they are the
subject matter of sociology and cultural anthropology, societies necessarily
have a social order but they may lack some sectors of the social order, such
as law, custom, manners, morality, or etiquette. We can imagine societies
so simple, as for instance the recently discovered Tasaday in the Philippines,
of which it may well be true that they do not have, say, a religion or a
morality, as they certainly do not have a legal order.
It is now fairly widely accepted that what we call "custom" and "law"
developed out of a simpler type of social order within which these distinctions
5-35
Of.
could not be drawn. This differentiation probably evolved together with the
growth of societies beyond the extended family and the consequent need to in-
clude in the social order rules regulating social intercourse between strangers.
Such regulation by undifferentiated mores has three serious defects. It is
uncertain because the compulsory social requirements are not clearly stated.
It is static , because there is no social institution which makes possible
deliberate and speedy changes in the social order. It is ineffective , be-
cause the social pressures which support the existing social order are inef-
ficient. The transition from a pre-legal to a legal social order occurs when
these three defects are remedied by the creation of certain social recognized
functions and roles: the function of determining what shall be part of the
social order backed by the strongest (legal) sanctions, a function performed
primarily by the legislator; the function of ascertaining whether a particular
person has acted contrary to the requirements of the compulsory social order,
a function primarily performed by the courts; and the function of dealing
appropriately with those who have been found to have acted in this manner,
a function primarily performed by the penal institutions. The mores of a
society encompass a legal order if and only if it has officials who ex officio
perform these tasks.
The basic judgments involved in law and custom are to the effect that
certain kinds of acts or certain particular acts are contrary to law or cus-
tom, or required or permitted by it. Practical reasoners will be expected
to take such judgments or supposed facts into account in their deliberations.
People's social training (their "socialization") will tend to bring it about
that such supposed facts are for them considerations relevant to decisions
5-36
to engage in behavior thus regulated. But such judgments have two important
characteristics which, from the point of view of practical reasoning, are
limitations. The first is that an action's being lawful or customary (or the
opposite) is so only within a given social order and at a given time. It is
not necessarily so within the same social order at another time, or within
another social order at the same or another time.
The second limitation is that such facts (that a course of action is con-
trary to law or custom) are not necessarily in themselves reasons against
such proscribed behavior. For what is law or custom in a given society at
a given time is determined by the social will, the will of its members as
whole—whatever the precise relation of the society's will to the will of
its members. In a political society that relation is in part consciously
determined by the political institutions. As we have seen in the discussion
Of Prisoners' Dilemma (see above, (4.5)), it is possible that no member wants
to do what the law (or custom) requires him to do, say, to pay taxes or serve
in the army, even if every member also wants every other member to do what
the law (custom) requires of him and wants there to be such a law. Even if
Rational Egoism is sound, such wants may be rational without its following
that these conventional requirements come to constitute reasons for the mem-
bers. Again, if moral principles constitute reasons, these conventional require-
ments need not necessarily be themselves reasons, for they may be unjust or
in other ways contrary to reason.
Thus, reasoning which leads to judgments, that certain courses of action
are unlawful or contrary to custom, in a certain sense hangs in the air. When
5-37
we have shown, by "subsumptive" reasoning (e.g., abortion is murder, or pris-
oners' consent is consent under duress), that a given course of action is
contrary to law or custom, we still have to show that such action is contrary
to reason. Most philosophers would agree that if law and custom did impose
at least prima facie moral obligations, then these conventional directives
would constitute at least prima facie reasons to act as they direct, but
there is considerable disagreement about whether they do impose such moral
obligations. It is therefore at least doubtful whether conventional reason-
ing is more than truncated practical reasoning. Only a satisfactory theory
of practical reasons would be able to settle this question.
Summing up our results so far, we found that there were two types of
truncated practical reasoning, technical and what I called "conventional."
Both argue from premises which, as experience shows, it will often be rational
for people to argue from. Both types of practical reasoning require to be
supplemented with reasoning from the domain of practical wisdom, i.e., the
theory of practical reasons. The main difference between technical and con-
ventional practical reasoning is that the former is by design reflexively bene-
ficial, the latter not. By following technical reasoning, one tends to gain
the benefits of attaining one's end. By following conventional reasoning, how-
ever, one tends, at best, to benefit others though, of course, general conformity
may be beneficial all round.
(10) Morality. We can now turn to what is for our purposes the most im-
portant part of a culture, its morality. It is also the most difficult and the
most contentious domain. In what follows I stress the rational and the social
5-38
aspect of morality. It seems to me that all the important clues point in this
direction. But though the views expounded below are widely shared, I cannot
pretend that they are not also widely rejected. The best that can be said for
this presentation is that no one who sets out to offer a coherent account
could hope to secure universal agreement.
Moral institutions resemble those of law and custom although, of course,
these roles are not occupied by professional specialists: everybody can and
over time does play all of them, as circumstances require. The role of moral
teacher is characterized by the fact that it involves not merely the teaching
of a skill, how to perform moral reasoning, but also inculcating the good will,
the readiness to perform these tasks up to the required standard of performance.
The central role, that of moral agent, is supported by several support roles.
There is the role of moral adviser who helps the agent with his moral reason-
ing. There are the roles of moral watchdog, moral accuser and judge, who
see to it that performers falling below the acceptable standards receive moral
condemnation. There is finally the role of defender or justifier, usually
but not necessarily the accused himself who can, by showing that he has per-
formed up to the required standard, abort moral censure. Lastly and most im-
portantly, there is the role of moral critic and reformer.
One of the important conclusions we can draw from a study of the way these
roles are performed is that, unlike one's Egoistic reasoning, one's moral
reasoning is a deep concern for others. In this respect it parallels one's
performance of conventional reasoning, especially legal. People are taught
to, not only how to, reason morally and to perform their practical tasks.
5^39
They are surrounded by performers of support roles who take an interest in
how well they perform and who censure them if they fall below standard. So
far, this is much like conventional reasoning and gives support to the widely
held view that a society's morality is simply a part of its mores. But such
a view overlooks the rational element in the way these roles are performed,
and in the language in which they express their reasoning and their conclu-
sions. When one arrives at the judgment that, say, the legislation in a
certain country of experimentation on aborted but live fetuses was morally
wrong, one is saying something stronger than when one claims that in a certain
country such legalized experimentation is unconstitutional. One is saying
not merely that such experimentation is contrary to a judgment arrived at by a
form of practical reasoning concerning which group members want others to
perform their theoretical and practical tasks adequately and support this
demand by social sanctions. It means, in addition, that there is reason for
them to demand this and to criticize anyone who performs poorly. It means,
further, that if the accused admits the judgment (that it really is wrong),
then he admits that the criticism was justified, that they are quite right
to judge that he ought not to have done what he did. None of this follows
from claiming that it really is illegal or unconstitutional. For if the law
is abolished, then such experimentation ipso facto ceases to be illegal even
if nothing else about such experimentation has changed. But in such a case
it could not also ipso facto cease to be morally wrong.
The second point of importance emerges from the study of the role of the
critic and reformer. A legal reformer may advocate changes in the law on two
sorts of grounds, internal or external, to remedy two sorts of fault. The
5-40
former are specifically legal flaws from some external point of view, such as
economy or justice. In the last case, the reformer is also a moral one, i.e.,
a reformer of the law on moral grounds—not of course a reformer of morality.
Now, the important thing to note is that there can be no such reform of moral ity
from an external point of view. It would be absurd to maintain that our
morality which frowns on slavery was in need of economic reform because moral-
ities which encourage slavery make for greater economic efficiency. I believe
this point strongly suggests that we think of moral reasoning as the highest
most inclusive form of practical reasoning.
This ties in with the way a moral reformer reasons. Even if he advocates
moral reforms, that is, reforms of currently accepted moral convictions, he
does not go outside morality. Such criticisms are always from within the
moral point of view. Moral thinking and reasoning is thus not in any way
limited or truncated. It is all of a piece from the most specific to the
most general principles. It seems to me that it is only because of this that
we can speak of our moral principles and precepts, such as "One ought not to
experiment on human subjects except with their consent" or "One ought not to
violate anyone's rights" or "One ought to treat others as one would have them
treat one" as moral convictions , that is, as normative convictions, as norma-
tive directives which can be sound or unsound, i.e., which pass or fail to
pass certain tests.
This leads to a third point. If moral reasoning is the highest type of
unlimited reasoning, that is, reasoning aiming at overall practical wisdom,
then its conclusions must be regarded as overriding those of all other such
5-41
types of reasoning, including those of Rational Egoism. This is, of course,
one of the widely held convictions we have about moral judgments, one very
strongly stressed in the moral philosophy of Kant. But if this is so, then
everyone must have adequate reason to do what morally speaking he ought to
do even if it runs counter to self-interest or desire or both. In other
words, the precepts and principles or moral reasoning themselves constitute
reasons for action, and the supreme ones at this.
It would seem to follow that the morality of a society belongs both in
its mores and in its practical wisdom. We must not, therefore, think of the
morality of a group as simply another sector of the mores, as is done, e.g.,
by the Legal Perspectives, when they construe what they call "positive moral-
ity" as parallel to law. A less misleading way to define the locus of
morality is to say that a society has a morality if it j_s_ a moral order, that
is, a rationally self-correcting social order in which this rational require-
ment is institutionalized in the role of the moral agent as also critic and
reformer, working as needed for a modification of the social order from the
moral point of view, that is, the point of view of one bent on creating a
social order which every member living under it has equal and adequate reason
to treat as overriding all other forms of practical reasoning including Egoistic
reasoning. The morality of a group thus overlaps with those parts of its
mores to which the members of the group give their backing because they believe
that they satisfy the requirements of morality. There is thus nothing surpris-
ing in the fact that we want to assign precepts such as "Commit no murder"
both to the mores and to morality, nor in the fact that in a society in which
morality is taken seriously, there is considerable disagreement among people
5-42
about what is right and wrong. The complexity of the issues and the tempta-
tion to pass off what is merely one's personal advantage as what is morally
right combine to make moral agreement extremely difficult. But if this
account is acceptable, it should at least be possible to recognize moral
reasoning and moral argument when it occurs, however uninformed or faulty
it may be.
Here we must note another important point. The morality of a group is
that part of the mores of the group which its members believe will pass a
certain test. We have already seen what this test is: to provide equal and
adequate reason for everyone in the community to accept these principles and
precepts of the mores as reasons to follow them, and as reasons overriding
all others. They must in other words be such that any rational person would
always follow them rather than those of any other type of reasoning with
which they might come in conflict. Precisely what this means will, of course,
depend on a satisfactory theory of practical reasons. But let us for argu-
ment's sake accept a limited theory of Rational Egoism, that is, one which
does not insist that self-interest and only self-interest constitutes a reason.
Then we can say that moral principles must be in the interest of everyone alike.
This is often interpreted to mean the same as being socially just. The morality
of the group would then be those elements of its mores which members of the
group believe satisfy the requirements of social justice.
We can now also say what, on this view, is the difference between a group's
moral principles and maxims, and sound or valid ones. Whereas the group's moral
principles are those its members take to be in the interest of everyone alike,
that is, just, the sound ones are those which actually are just. Again, only
5-43
a sound theory of justice can give us knowledge of what this criterion comes
to. The recent work by John Rawls does show, however, that progress here is
not impossible.
It remains to say, briefly, what it is for an individual to have a
morality. The sense in which he can have a morality all by himself is that
in which he can have a grasp of moral reasoning, its nature and function,
so that he can in any moral order in which he may find himself play the
various moral roles including that most important role of moral reformer.
But if my account is correct, then an isolated individual who has no contact
with others, such as, let us say, the last survivor on a spaceship, would not
be able to employ his moral reasoning to any purpose, except perhaps to main-
tain those excellences, such as self-control, which are presuppositions of
adequate moral performance, just in case he does later rejoin a human group.
A second consequence is that there are indeed certain things the moral
agent can take with him to other societies, namely, those which are indepen-
dent of social institutions, such as the moral principles which govern what
one person must not or may do to_ another (e.g., killing, injuring), and
what he must or need not do for another (e.g., attending to a sick person
if he is a physician), but of course even in this realm he may find the
other society to have quite different principles and precepts, and he will
have to act in relation to that society in the same way as would a member
who wants to reform it. He cannot simply , and without consideration of the
accepted morality, follow his own moral convictions.
5-44
The third is that it is conceivable that in a group without a social
order, such as Hobbes state of nature, he will want to use his moral prin-
ciples, at least those mentioned in the previous paragraph, which do not
depend on the institutions of a social order. What fate he may there suffer
if he acts in this way will of course depend on what these people are like.
There is no reason to think that his morality will then lead to a fulfilling
life. Nor can anyone in these circumstances use the moral terms, "right" and
"wrong" in the way we now do. It is as if, after poison gas has killed most
of the inhabitants of a country, but in some cities the traffic lights are
still functioning automatically, the few motorists left asked themselves
whether it would be wrong for them to cross when the signals are red.
IV. Guidelines
(11) Identifying Ethical Principles, Reasons, Objections, Arguments.
If the account of morality given in Section III is correct, then we must
distinguish between what a person thinks or sincerely says his moral prin-
ciples etc., are and what they really are. He may. make three sorts of mis-
takes: what he offers are not principles, etc. (but, say, only rules of thumb),
they are not his (he is only paying lip service to them), and they are not
moral (but prudential or technical). We need not, in this context, worry
about the first two mistakes, but the third is relevant and a real possibi-
lity. Even if people can argue validly from moral principles, they may have
quite mistaken views about what they are doing when they do it, and popular
misconceptions of morality may seriously blind them to what is implied in
their own judgments and practices. On the view outlined in Section III, moral
5-45
principles, precepts, maxims, reasons, objections, and arguments are those
which are employed in moral reasoning. Moral reasoning is a form of prac-
tical reasoning which is practical reasoning from the moral point of view.
One adopts that point of view when one considers one's social order as
grounded by a set of supreme guiding principles for all its members, and
thinks of these members as rational agents willing to be guided by such
principles as long as they have reasons for doing so as good as have any
of the other members, that is, as long as the principles are socially just.
The second, even more important and difficult question raised by the
Commission is how we distinguish between "valid" and "invalid" (or sound/
unsound) ethical principles, and so on. The answer has already been given
before. Sound or valid moral principles, are those which not merely purport
to provide everyone with equal and adequate reason to follow them, but
really do so. Whatever is the correct answer to the question of what justice
is, it will undoubtedly involve respecting everyone's moral rights. A con-
cern with the rights of individuals affected by proposed courses of actions
or policies will therefore be a strong indicator of the reasoner's reason-
ing from the moral point of view.
(12) A simple example will perhaps help to make this more concrete.
Dr. William Smith, in the Third Academy Forum of the National Academy of
Sciences, entitled Experiments and Research with Humans: Values in Conflict ,
1975, p. 90, gives the following hypothetical case.
Haym's Syndrome is a disease of the joints that begins in
early childhood and becomes progressively more debilitating
through early adulthood, severely impairing joint function
in about half the cases. The death rate among Haym's Syn-
drome victims is about 10 percent before age eighteen, in-
5-46
creasing throughout adulthood to about 40 percent by age
forty.
All groups within the American population are at equal
risk for this disease.
There is a conventional treatment that may bring symp-
tom relief, but there is no known cure.
Several competent investigators concluded after six
months of animal studies that a new drug, NAS-18, if ad-
ministered continuously in high dosages could well arrest
the progress of the disease.
Their animal trials, however, have also indicated several
discomforting although not serious side effects, including
diarrhea and insomnia.
The investigators decided on a two-year trial at their
Haym's Syndrome Clinic at University Hospital. In their
NIH grant application the trial was stated to have the
dual purpose of measuring the toxicity and the beneficial
effects of NAS-18 and weighing them against each other.
Forty potential subjects were selected. Thirty were
outpatients at the Haym's Syndrome Clinic. Of these
thirty, ten were children ranging in age from five to
seventeen.
One of the investigators was the staff physician at the
local juvenile training school. Ten adolescents age ten
to fifteen at the school had Haym's Syndrome and were
therefore selected as the remaining subjects.
At their next clinic visit, each of the thirty out-
patients -- twenty adults and a parent of each of the
ten child subjects -- was asked to participate in the
research project. One of the investigators carefully
explained the following to each person:
- That NAS-18 was a new drug that had shown pro-
mising results in previous tests. It might
possibly be a major breakthrough in the cam-
paign against Haym's Syndrome, and it might
lessen the subject's pain. No one, however,
could be sure of this until further research
was undertaken.
- That the procedure simply involved taking
pills on a regular basis and that their pro-
gress would be carefully monitored.
- That previous research had indicated some
chance of developing the known side effects
described earlier but that there could be
other, unknown side effects.
The patients and the parents of the child subjects
were then given a chance to ask questions and asked to
sign the standard consent form used by University
Hospital. [That form follows the statement of the
case.]
5-47
Parental consent was obtained for all but three of the
ten training school subjects whose parents could not be
located. The school's chief administrator's consent also
was obtained. The subjects then met as a group with the
investigator and were told that they were being asked to
participate in a drug test that had been approved by the
administrator. They were given the same information as
the clinic patients, but they were not asked to sign a
consent form. They were told that they could withdraw
from the experiment at any time.
The research proceeded according to design. Half the
adults and half the children were given NAS-18, the
other half a placebo. Dosages for all subjects were in-
creased periodically during the first year of the research.
Many of the subjects seemed to benefit substantially
from the administration of NAS-18 — especially the
children whose disease had not progressed to its most
debilitating stages. AH subjects on NAS-18 experienced
less pain.
However, at a dosage where NAS-18 appeared to slow the
progress of the disease, five subjects developed peri-
pheral neuropathy, three so severe they had trouble walk-
ing. The investigators took all five subjects off NAS-18
and reduced the dosage administered to the other fifteen.
Neither the remaining fifteen NAS-18 subjects nor the
twenty control subjects were informed of the neuropathy.
Four months later five more subjects developed peri-
pheral neuropathy. They then were taken off the drug.
Near the end of the trial , many of the subjects whose
symptoms had been most relieved told the investigators
they wanted to continue NAS-18 treatments. Expecting
an affirmative response, they were surprised and disap-
pointed when told that further trials were necessary
before NAS-18 could be approved as accepted treatment
for Haym's Syndrome and that they could not continue on
the drug after the end of the two-year trial .
At the conclusion of the experiment, all patients were
taken off NAS-18, told that further research would be
undertaken but not at University Hospital, and thanked
them for their cooperation and help in finding what ap-
peared to be a promising therapy for Haym's Syndrome.
Six patients who continued to show abnormal nerve
conduction and signs of peripheral neuropathy continued
their visits to the clinic.
In the discussion of the case, Dr. Smith then makes a number of claims
about what should and should not have been done in this case. He says, for
instance: "This experiment should never have been approved by the University's
-48
Research Review Committee or funded by the National Institutes of Health.
The researchers gave children a drug that was never before tested on humans.
There was no need for this, and no justification. Children should not have
been used until the toxicity of NAS-18 had first been tested on adults,
especially since the experiment could, at best, have benefited the subjects
only temporarily. Even captive children were used, and used purely for the
convenience of the investigators. Institutionalized children should not have
been placed at risk when other subjects were available. Three children were
used without the consent of even their parents. No child, at least one whose
natural or adoptive parents were not available to give informed consent,
should be the subject of a non-therapeutic experiment. None of the children
were themselves asked to consent. I submit that no child over seven should
be placed at risk in a nontherapeutic experiment without his or her approval."
(p. 92)
Several of the statements amount to formulations of principles or are
straightforward applications of such principles: "Children should not be
used until the toxicity of NAS-18 had first been tested on adults, especially
when the experiments can benefit the subjects only temporarily." "Institu-
tionalized children should not be placed at risk when other subjects are
available." "No child over seven should be placed at risk in a nontherapeutic
experiment without his or her approval."
(13) Moral and Technical Reasoning. These are low-level principles
or precepts which may or may not be intended and defended as ethical ones.
It would depend how Dr. White would argue for and meet objections to them.
5-49
If he had argued, for instance, that there should have been more extensive
testing on animals first, that it was wasteful and inexpedient to use scarce
and valuable child subjects to test the toxicity of NAS-18 when it could be
done more cheaply and without unfavorable public reaction on animals, this
would be a technical argument to the effect that those interested in ensur-
ing continued research opportunities on children were going about it in a
counterproductive as well as inefficient way. If, however, he had gone on
to say that the experiment on children was morally unjustified because it
unjustifiably and unnecessarily put children at risk, when adults or animals
could have been used, then presumably he intended it as a moral principle,
one not merely defensible but required from the moral point of view.
(14) Moral and Legal Reasoning. In a later comment, Dr. Smith says,
"I seriously question whether any physician should be permitted to conduct
a nontherapeutic experiment on his or her own patients. Certainly a research
subject's health interest should always be protected by a third party not in-
volved in the research." (p. 94) The underlying principle is, of course,
"No physician ought to be permitted ..." Such a principle would presumably
be defended on what has recently been called "exclusionary reasons." 29 The
reason for advocating this principle is not, presumably, that researchers
are less capable of recognizing the health interests of their subjects, or
less moral, than other people, but that their research interests are in con-
flict with their subjects' health interests, and that this would place a con-
siderable burden on them in performing their theoretical and practical tasks.
The principle, in other words, excludes the researchers, not because it is
in itself immoral to perform these experiments, but because it is in itself
5-50
immoral to perform these experiments, but because they cannot be relied
on to perform the relevant theoretical and practical tasks involved in
reasoning about their subjects' health needs, as confidently as nonparti-
cipant physicians could be. Thus, when it comes to the secondary question
of who should be legally permitted to perform these experiments, perhaps
who should be called upon to decide who should be permitted - not whether
it would be wrong to perform such experiments - then this exclusionary
reason is in itself a moral reason. It is important to note that this
secondary question, 'Who should be allowed ... ?"is not itself a legal ques-
tion, but a moral one. The parallel legal question would be, "Who is (by
law) permitted ... ?"
on
(15) Utilitarian Reasoning. In another example in the same volume,
Dr. Francis D. Moore relates the story of how 250 years ago, a Dr. Zabdiel
Bolston, encouraged by the Reverend Cotton Mather, conducted a research
experiment without any preliminary animal experiments whatever. He inocu-
lated persons with the virus of smallpox during a smallpox epidemic. The
results: while the natural disease produced a 15 percent mortality, the
individuals who received the inoculation suffered a 2 percent mortality.
Dr. Moore does not comment directly on this case, but he appears to think
that the risks were well worth taking, in view of the developments for
which it paved the way. This is an argument based on utilitarian considera-
tions. The benefits gained by future generations are thought to outweigh
the harm and the risks undergone by those subjected to the experiment. Some
utilitarian moralists would no doubt be inclined to treat this as a moral
31
argument, but others who lean towards so-called rule-utilitarianism would
5-51
••eject it. Be that as it may, that such a utilitarian is arguing morally
would show itself in the way i"n which, if he accepts the principle that very
great future gains justify present risks and losses, he is concerned to re-
duce these present risks to a minimum and to distribute them fairly.
(16) Political Reasoning. Suppose now that one of the discussants is
a Congressman, and that he argues against a legislative proposal requiring
all hospitals to set up Hospital Research Review Committees on which the
general public is strongly represented, on the grounds that he would lose
the next election if he supported such legislation, or that such legislation
would never get through Congress even if he himself voted for it, or that if
he succeeded in mobilizing enough support for this, he would in return have
to give support to other measures he disapproved of. These would not be
moral but political arguments, and they, too, would be technical arguments,
as explained above.
5-52
NOTES
1, P. 3: I do not here distinguish between 'ethical' and 'moral'. But I do
distinguish between 'ethics' and 'morality', Morality contrasts either
with immorality or with law, religion, manners, and the like. Ethics
does not have either of these opposites, A history of ethics is a his-
tory moral (or ethical) theory-- it is neither a history or morality
from its emergence out of customs, nor a history of morals, what peo-
ple thought right and wrong and the extent to which they lived up to
their convictions.
2, G,E. Moore, Principia Ethica , Cambridge University Press.
3, W.D, Ross, The Right and the Good , Clarendon Press, 1930.
4, C.L. Stevenson, Ethics and Language , Yale University Press, 1930.
5, R,M. Hare, The Language of Morals , Oxford University Press, 1952;
Freedom and Reason . Oxford University Press, 1965.
6, S.E, Toulmin, The Place of Reason in Ethics, Cambridge University Press,
1950.
7, See note 5.
8, On this topic the literature is extensive. I mention three widely read
books: David P. Gauthier, Pr actical Reasoning , Clarendon Press, 1963,
Roy Edgely, Reason in Theory and Practice , Hutchinson University Library,
1969, David A.J. Richards, A Theory of Reason for Actio n, Oxford University
Press, 1971.
9, Patrick Nowel 1 -Smith, Ethics , Pegu in Books, 1954.
10. See for instance, C.H. Whitely, 'On Defining "Moral"', Analysis , 1959-60.
11. See e.g., D.W. Falk, 'Morality, Self, and Others', in Castaneda and Nakhni-
kian, ed, Morality and the Language of Conduct , Wayne State University Press,
1963.
12. See e.g., W.K. Frankena, 'The Concept of Morality', University of Colorado
Studies , 1967.
13. See e.g., Alaskair Maclntyre, 'What Morality is not', Philosophy 1957; also
C.H. Whitely, op, cit., and R.M. Hare, op. cit.
14. J.P, Sarte, Existentialism is a Humanism . Also R.H. Hare, op. cit., and
'Universalizability' , Proc. Arist. Soc, 1954-5.
15. R.M. Hare, The Language of Morals , pp. 13 f. and 20 f.
5-53
16. Cf. W.K. Frankena, op. cit,
17. Cf. Philippa Foot, 'Morality as a System of Hypothetical Imperatives',
Phil. Review , July 1972,
18. The sense of 'rational' as opposed to 'non-rational' is here irrelevant,
since in this context, rationality in this sense is presupposed. He are
not talking about animals which are non-rational beings, not even about
fetuses which are merely potentialy rational beings, but only about ac-
tually rational beings, those who already have the ability to engage in
reasoning. What we are interested in is the distinction, within that
class of beings, between those whose performance comes up to a certain
standard and those whose performance falls below.
19. Such a view was probably held by C.L. Stevenson. See Ethics and Language .
Also by David Falk, e.g., 'Action-guiding Reasons', Journal of Philosophy ,
Nov., 1973.
19a, See e.g., R. Duncan Luce and Howard Raiffa, Games and Decisions , John
Wiley & Sons, 1965, pp. 94-97.
20. A brief discussion of these principles is contained in John Rawls, A Theory
of Justice , pp. 408-416.
21. This is, of course, Hume's view or at any rate the view commonly attributed
to Hume, Until recently, it was also the view commonly attributed to Aris-
totle. For a different view, see the recent work of John Cooper, Reason
and Human Good in Ar istotle , Harvard University Press, 1975.
22. Cf. Luce and Raiffa, op. cit.
23. See David Lyons, In the Interest of the Governed , Oxford University Press,
1973.
24. The case is taken from The Hastings Center Report , Vol. 5, No. 1, February,
1975, p. 49.
25, Such technical syllogisms must be distinguished from what is often called
the Practical Syllogism, which are always first-person arguments and whose
conclusion is the reasoner's action or his setting himself to act. An exam-
ple would be: (i) I am set to cure this patient, (ii) If I am to cure this
patient, I must insert a tube, (iii) So I am setting myself to insert a
tube. We can ignore this form of reasoning since whether or not is is a valid
form of reasoning it will occur as part of the performance of the practical
task in all practical reasoning, and will not therefore help us distinguish
moral from other kinds,
26, See, e,g,, H.L.A. Hart, The Concept of Law , Clarendon Press, 1961, pp. 89-96.
27. See, e.g., Richard A. Wasserstrom, 'The Obligation to Obey the Law', Uni-
versity of California Los Angeles Law Review , Vol. 10, May, 1963. See
also Hart, op. cit., Chs. V, VIII, IX.
5-54
28. See e.g., John Austin, The Province of Jurisprudence Determined , Weidenfeld
and Nicholson, 1954. Lecture V.
29. See J. Raz, 'Reasons for Action, Decisions and Norms,' Mind , Vol. LXXXIV,
No. 336, Oct. 1975.
30. Third Academy Forum, Experiments and Research with Humans, pp. 16-19.
31. For an excellent if somewhat difficult discussion see David Lyons, The
Forms and Limits of Utilitarianism . Clarendon Press, 1965. A brief his-
torical survey of the main forms of utilitarianism are contained in Anthony
Quinton, Utilitarian Ethics , Macmillan, 1973. A stimulating confrontation
between a utilitarian and an antiutil itarian can be found in J.J.C. Smart
and Bernard Williams, Utilitarianism, For and Against , Cambridge University
Press, 1973.
5-55
6
DISTRIBUTIVE JUSTICE AND MORALLY RELEVANT DIFFERENCES
Tom Beauchamp, Ph.D.
I. The Concepts of Justice and Distributive Justice
The Meaning and Types of Justice . Sidgwick, Rawls, and others
have argued forcefully that our basic notion of justice is more akin
to the notion of fairness than to most any other moral notion. While
they are right to insist on the close conceptual connections between
these terms, perhaps the single word most closely linked to the general
meaning of "justice" is "desert." One has acted justly towards a
person when that person has been given what he is due or owed, and
therefore has been given what he deserves and can legitimately claim.
It may be that a person deserves to be awarded a prize, for example,
in which case justice has been done when that person receives the prize.
What persons deserve and hence can legitimately claim is based on cer-
tain relevant properties which they possess. Because a person possesses
the property of being a law-breaker, or of otherwise wrongly treating
others, we are justified in allocating an appropriate punishment. But
it is wrong, as a matter of justice, to allocate a punishment or re-
ward if the person does not possess the relevant property. It is wrong
or unjust of a man to punish his dog for knocking a child down when
in fact it was the man's negligence which led to the incident. Similarly,
it is unjust to reward a superior for the work of his subordinates when
it was not his guidance which led to rewardable productivity.
The expression "distributive justice" refers to the proper dis-
tribution of social benefits and burdens. Paying taxes and serving on
juries are distributed burdens, while welfare checks and foundation
grants are distributed benefits. Recent literature on distributive
6-1
justice has tended to focus on considerations of fair economic distri-
bution, especially unjust distributions which occur in the form of
inequalities of income between different classes of persons and unfair
tax burdens on certain classes. But clearly there are many problems
of distributive justice.
Comparative and Non-comparative Justice . It is also crucial to
the idea of justice that deserts be comparative or comparable. Justice
is said to be comparative when what one person deserves is measured by
balancing the competing claims of other persons against his claims.
Here the social condition of others affects how much an individual is
due. Justice is non-comparative , in contrast, when desert is judged
by a standard which is independent of the claims of others (e.g. an
innocent man never deserves punishment). This paper deals exclusively
with comparative justice.
Scarcity and Distrib utive Justice . It is noteworthy that distrib-
utive justice is a notion which applies only to the distribution of
scarce benefits, where there is some competition for the benefits. If
there are plenty of fish in a river for everyone to have as many as he
can catch, we do not establish patterns of distribution. It is only
when we are worried that the fish supply will be exhausted or that future
fishermen will be unfairly affected by present fishing that we set limits
to the number of fish they may catch. There are, of course, various
schemes which could be devised for the distribution; but that is not the
present point. The point is that there are no problems of distributive
justice and no need of principles of distributive justice until some
measure of scarcity exists. Even when burdens rather than benefits are
being allocated, there is competition for the least disadvantageous
distribution.
6-2
David Hume pointed out that we have developed the concept of
justice in order to handle problems of conflicting interest when
claims are being pressed by conflicting parties. As he put it, there
would be no point to having rules of justice unless society were
composed of persons of limited sympathy in the competition for scarce
resources. The rules of justice serve to strike a balance between
those conflicting interests and claims that repeatedly occur in society.
This shows a close link between the lawful society and the just society,
since law and morality are our explicit tools for balancing conflicting
1
claims. Nonetheless, the law may be unjust; and there may be many
rules of justice which are not connected to the law or to legal enforce-
ment. There seems to be no escape, then, from the conclusion that the
notion of comparative justice presupposes parties with conflicting claims
who attempt to justify their claims by appeal to basic moral rules.
Just Procedures and Just Results . There also exists an important
distinction between just procedures and just results . The term "dis-
tribution" may refer to the process or procedure of distributing, or
it may refer to the product or result of some system of distribution.
Ideally, of course, it is preferable to have both just procedures and
1
"Justice" is sometimes used as a virtual synonym of "lawful," and some-
times even of "good." Thus, it is often said that the just society is
the lawful society or the good society. While it may be true that the
just society is both lawful and good, it cannot plausibly be claimed
that this characterization captures the meaning of "justice" very well.
6-3
just results, but it is not always possible to have both. For ex-
ample, one might achieve a just result in distributing health care,
but one might use an unjust procedure, such as illegal taxation of
certain groups, in order to achieve it. One might also have a just
procedure, e.g. a lay-professional interdisciplinary ethics committee
in a hospital which selects patients for dialysis, and yet reach un-
just results (since, say, certain information was not available or a
member of the committee was prejudiced, etc.). In discussions of dis-
tributive justice it is often the system which is in question, and yet
those who are criticizing the system may do so by pointing to unjust
results. It is important to be clear, then, on whether it is the pro-
cedures or the results which are under consideration.
Formal and Material Prin ciples of Justice . Justice in the sense
of comparative desert has been analyzed in different ways in rival
theories. But common to all theories of justice is a rather minimal
principle which is traditionally attributed to Aristotle: equals ought
to be treated equally and unequals unequally. This elementary principle
is referred to as the principle of formal justice, or sometimes as the
principle of formal equality. It is said to be formal because it states
no particular respects in which equals ought to be treated the same.
Effectively it says that no matter what respects are under consideration,
if persons are equal in those respects, then they must be treated equally.
More fully stated in negative form, the principle says that no person
should be treated unequally, despite all differences with other persons,
until such time as it has been shown that there is a difference between
them relevant to the treatment at stake. In a more positive form, the
6-4
principle says that individuals who are equal in the relevant respects
must be treated equally, while individuals who are unequal in the
relevant respects whould be treated differently in proportion to the
differences.
The problem with the formal principle is notoriously more in its
aloofness than in any deficiency of content. That equals ought to be
treated equally, by law and elsewhere, is not likely to stir disagree-
ment. But who is equal and who unequal? Presumably all citizens should
have equal political rights, equal access to public services, and should
receive equal treatment under the law. But almost all would allow that
distinctions based on experience, deprivation, merit, and position do
sometimes introduce criteria justifying differential treatment.
Any plausible theory of justice, then, must explain the respects
in terms of which people should be equally treated, which involves
specification of relevant properties. And it seems clear enough that
not just any proposed criteria are morally fair. For example, if it
is judged a good reason for not interviewing women for jobs that they
make male interviewers nervous, then this introduces a proposed relevant
difference. Yet this difference scarcely seems compatible with justice,
since it allows a blatant injustice based on a morally irrelevant property.
The really difficult questions about justice arise when we try to specify
the relevant respects in terms of which people are to be treated equally.
Any principles which so inform us are said to be principles of material
justice, because they put material content into the theory. Examination
of some major material principles of distributive justice is the next
order of business.
6-5
II. Principles and Theories of Distributive Justice
In the philosophy of distributive justice there are a few widely
discussed, and almost as widely accepted, material principles of dis-
tributive justice. Each principle mentions a relevant property on the
basis of which burdens and benefits should be distributed. That is,
each principle asserts a standard of relevance for purposes of dis-
tribution. The standard is normally, though not necessarily, a property
which persons possess or fail to possess. What makes each principle a
plausible candidate is the plausible relevance of the property it isolates.
The following is a fairly standard list of the major candidates for the
position of valid principles of distributive justice (though still differ-
2
ent and longer lists have been proposed ) :
1. To each person an equal share
2. To each person according to individual need
3. To each person according to individual effort
A. To each person according to societal contribution
5. To each person according to merit (individual ability)
There is no a priori barrier to acceptance of more than one of these
principles, and some theories of justice accept all five as valid. Most
societies use several of them, applying different principles of distri-
bution in different contexts. In the United States, for example, unemploy-
ment and welfare payments are distributed on the basis of need (and to
some extent on the basis of previous length of employment); jobs and pro-
motions are in many sectors awarded (distributed) on the basis of demon-
strated achievement and merit; the higher incomes of wealthy professionals
2
cf . , e.g., Nicholas Rescher, Distributive Justice (Indianapolis: Bobbs-
Merrill, 1966), Chapter 4.
6-6
are allowed (distributed) on the grounds of superior effort or merit
or social contribution (or perhaps all three); and, at least theo-
retically, the opportunity for elementary and secondary education is
distributed equally to all citizens.
Theories of distributive justice are commonly developed by em-
phasizing and elaborating one or more of the material principles of
distributive justice, perhaps in conjunction with other moral principles.
Thus, Egalitarian theories emphasize equality; Marxist theories emphasize
need; Capitalist theories emphasize contribution and merit; Utilitarian
theories emphasize a mixed use of such criteria so that public and private
utility are maximized, etc. The viability of any such theory of justice
is determined by the quality of its moral argument to the conclusion that
some one or more selected material principles ought to be given priority
(or perhaps even exclusive consideration) over the others.
Obviously consideration cannot here be given to the detailed nature
either of principles or of theories of distributive justice. Nonetheless,
it is important to see how from s^uch meager and abstract beginnings as
"the principle of " both relevant properties and public
policies based on justice can be developed. Consider, as an example,
the development of policies based on the principle of need. The principle
of need declares that distribution is just when it is based on need.
But how are we to understand the notion of a need? Certainly the term
is subject to different interpretations, and one must settle on a
meaning before proceeding further. In general, let us suppose, to say
that a person needs something is to say that without it he will be harmed
(or at least detrimentally affected). We can fill this basic idea out
6-7
by calling on the formal principle of justice, which requires us to
say that people of equal need should be treated equally in regard to
the satisfaction of these needs, while those who have unequal needs
should be treated unequally. There are instructive examples of our
acting on this principle which show its immediate relevance to contexts
of distribution. In hospital wards people are given equal amounts of
blood when they need equal amounts and unequal amounts when the amount
needed is unequal. However, this analysis of needs does not take us
far, since we are not required to distribute equally for all needs —
such as needs for pets, athletic equipment, nightgowns, peaches, etc.
(unless one is prepared to defend a radical form of egalitarianism) .
Presumably we are interested only in fundamental needs, so we ought to
look further at this notion.
To say that someone has a "fundamental need" for something is to
say that he will be harmed or detrimentally affected in a fundamental
way if he does not have that thing. Examples of fundamental harms would
be malnutrition, serious bodily injury, and the withholding of critical
information. Without certain forms of food, health care, and education,
these harms may befall us; hence we say we have a fundamental need for
certain foods, health care facilities, and shelter. Generally these are
needs which are necessary for survival, or more accurately for survival
in a state of existence which is itself better than non-survival.
Now if_ one accepts a theory of distributive justice which permits
appeal to the principle of need, then the more one refines the notion
of needs, as we have done in preliminary fashion above, the closer one
moves toward the relevant properties necessary for the formulation of
6-8
a public policy position. For example, if there exists a fundamental
need for health care facilities, then it would have to be decided
which needs are fundamental and which are not in order to develop a
national health care policy. Anyone who had such needs would have
the relevant properties. While this sort of theoretical refinement
cannot be carried out here, it is of vital importance to notice the
role of the first step in the argument — the acceptance of the principle
of need as a valid material principle of justice. If one rejects
rather than accepts the principle of need (while accepting, say, only
contribution and merit), then one could judge such refinements in-
applicable as a matter of justice and would be opposed in principle
to the refinements and applications to public policy mentioned above.
All public policies based on distributive justice derive ultimately
from the acceptance of one or more material principles of distributive
justice and from some procedure for refining them.
III. "Relevant" Respects and Classes of Subjects
The notion of relevant respects or relevant properties is a particu-
larly vexatious one, and this section is devoted exclusively to its
elucidation. I begin with a recapitulation of the previous two sections,
as they have introduced the problem of relevant respects.
The formal principle of justice requires that individuals who are
equal in the relevant respects be treated equally. Material principles
of justice then specify what the relevant respects are, thereby ex-
eluding other respects as irrelevant to the treatment under consideration.
A simple example is provided by a large company which offers free in-
house legal services for employees. Suppose the company distributes
6-9
jobs on the basis of demonstrated ability, salaries on the basis of
contribution, and free legal services on the basis of need. These
three material principles fix relevant differences and also fix rele-
vant similarities. Differences in pay, for example, are determined
by differences in contribution (the relevant differences in this
context), while similarities in pay are established by similarities
in contribution (the relevant similarities) . Obviously the same can
be said about the allocation of jobs and legal services.
Moreover, further detailed specification of the criteria of ability
(e.g., a certificate from a training school), of contribution (e.g., a
salesman's total monthly sales), and of need (e.g., any legal service
without which a person would be detrimentally affected) serves to fill
out the precise nature of the relevant respects. These relevant respects
can be correctly or incorrectly fixed. They are correctly fixed when
supported by moral principles and incorrectly fixed when unsupported by
moral principles (or when supported by incorrect "moral" principles) ,
as we shall now see.
Established Relevance and Justified Relevance . However obvious
the preceding may appear, there are theoretical difficulties in ex-
plicating the notion of relevant respects, as well as practical problems
in the development of public policies. In some contexts relevant re-
spects are firmly established, perhaps by tradition and perhaps by
moral principle. Here it would be generally (though not always) in-
appropriate to challenge the established relevant respects by attempting
to substitute others. For example, trophies are awarded (distributed) at
the end of tennis tournaments on the basis of achievement; and how
6-10
achievement is to be determined is firmly set by the tradition-bound
rules of tournament tennis. Similarly, as a moral matter, prison terms
are not distributed to those who are not found guilty of crimes, since
as a firm matter of law and morality, guilt is relevant to conviction.
However, in many controversial contexts it is morally appropriate
either to institute a policy which establishes relevant respects where
none has previously been very firmly established or to develop a new
policy which revises standard or proposed "relevant" respects. If a
person is chosen to be an ambassador to a foreign nation merely on the
basis of wealth, party affiliation, and loyalty to the chief executive,
it is arguably the case that these operative (and perhaps even traditionally
entrenched) "relevant" properties are from the point of view of justice
arbitrary and irrelevant. [Here it would be argued that: "You ought as
a matter of justice to shift your operative set of 'relevant' properties
to the right set of relevant properties, which would include linguistic
facility, knowledge of the country, administrative experience, past
contribution, etc."] Similarly, if football scholarships to a university
are awarded simply on the basis of ability as a football player and with-
out reference to scholastic ability, it is arguably the case that the
"irrelevant" property of scholastic ability ought as a matter of justice
to be recognized as relevant. [Under the assumption that such resources
are scarce and highly competitive, it would be argued here that; "You
ought as a matter of justice to modify your operative set of 'relevant'
properties to include scholastic ability."]
In these examples it is being argued that from the point of view of
justice certain properties accepted as "relevant" are actually irrelevant
6-11
and that certain properties presumed "irrelevant" are actually relevant.
The Role of Argument and Decision . It is widely assumed that rele-
vant properties are fixed independently of both moral argument and human
decision, which establish them as either relevant or irrelevant. Just as
the applicable rules determining the eventual winner of a tennis tourna-
ment are valid whether we know the rules or not, so we are tempted to
think when some moral controversy arises that relevant properties are
similarly fixed independently of human awareness of them. There is some-
thing almost certainly right, but something just as certainly misleading
about this thesis. What is right is that basic principles of morality
control the relevance and irrelevance of properties. If basic moral
principles are not arbitrary and not changeable merely by individual
human fiat (an assumption of moral philosophy which is here taken for
granted) , then the properties which such principles determine to be
relevant are also neither arbitrary nor mere matters of individual pre-
ference. Any decision, then, will be one with attached moral constraints.
This important point requires further explanation.
Consider the issue of experimentation on adult human subjects. How
shall this burden be distributed? The formal principle of justice de-
clares that we must treat everyone who is alike in the relevant respects.
But which respects are here relevant? Suppose the members of one class
of adult persons have consented to participation, are informed regarding
the experiments and their risks, and have not been coerced in giving
their consent. Suppose, however, that the members of another class of
persons either have not consented, or are not informed, or have been
coerced. We would all agree that consent, understanding, and non-coercion
are clear cases of non-arbitrary relevant differences (whether or not
6-12
these properties are sufficient conditions of justifie d experimentation).
But why are these respects relevant, while properties such as tallness,
handsomeness, and I.Q. are not relevant? The answer is that their rele-
vance is non-arbitrarily fixed by moral principles. Non-coerced, informed
consent is relevant because basic moral principles such as respect for
persons and human autonomy determine relevance. Stated negatively, if
consent, understanding, and non-coercion were not present, then (coerced)
experimentation would violate basic moral principles protective of human
autonomy and respect for persons. No parallel moral constraints require
consideration of tallness, handsomeness, and I.Q. This also explains
why we often use the language of morally relevant properties and morally
irrelevant properties.
We insist that the choice of such properties is not morally arbitrary
or merely a matter of subjective preference, precisely because their
selection is backed by moral principles.
On the other hand, there are occasions where moral principles do not
in any clear fashion determine relevant properties. Usually this occurs
not because moral considerations are unimportant, but rather because
there are conflicting moral demands where no single moral principle is
determinative. In such cases a moral decision concerning the weight of
competing moral claims is required, and this decision in turn fixes the
acceptable relevant properties. whether members of minority groups
formerly discriminated against should be given preferential consideration
in hiring is one such issue with important policy implications. Whether
18-year olds should be allowed to vote is another. A particularly
striking example is found in the current controversy whether a transsexual
6-13
female should be allowed to compete in women's tennis tournaments.
Presumably the only relevant property governing qualification for
such a tournament (assuming the requisite tennis ability) is that of
being a woman. However, possession of the property of transsexuality
in this context seems to throw into open question what the appropriate
criteria are for being a woman. We are then invited to consider whether
chromosome structure is relevant or irrelevant. This issue at first
appears to be either an empirical or a conceptual dispute concerning
necessary and sufficient conditions of being a woman. More plausibly,
it is a moral problem of justice. On the one hand, the question is
whether it is fair to non-transsexual women to allow transsexuals to
compete; and, on the other hand, it is asked whether it is fair to
transsexuals that they be excluded. Such an issue will almost surely
be decided primarily on a moral rather than a conceptual basis. It
will be decided by considering the weight of moral arguments on each
side, and in the end the relevance or irrelevance of chromosome structure
and of other proposed criteria will be decided by reference to moral
standards of fairness.
A great many moral problems and public policy decisions take precisely
this dilemmatic form. There are powerful moral reasons in such cases
for accepting two or more sets of different and competing properties as
equally relevant, even though only one can be adopted and put in service.
It would be convenient if relevant properties were always fixed in the
way tennis rules and criteria of informed consent are fixed, but often
they are not settled in these ways and must be explicitly fixed by moral
deliberation and decision. Usually in such cases it is neither unreasonable
6-14
nor unfair if the final decision favors either of two or more competing
positions.
Classes of Subjects . Classes of subjects which are the objects
of moral concern should be grouped for ethical purposes only by the
relevant properties which they share in common, though there is often
a temptation to group them in terms of convenient but irrelevant properties.
Suppose a statute governing jury duty excuses all men but excuses no
women, on grounds that there are many more men in the working population
than women. Here justice requires us to say that an undue burden is
placed on women and an undeserved privilege is granted to men. Being
male or female is irrelevant, and so is the sexual grouping. If the
relevant property excusing jury duty is employment, then employed men
and employed women should be excused without regard to sex.
Or consider again the complicated matter of human experimentation,
in this case with special reference to prisoners. For simplicity suppose
there are two and only two classes of prisoners: (1) those whose re-
strictive environments are as non-coercive as normal non-prison environ-
ments and (2) those whose restrictive environments are inherently coercive.
Suppose further that laws governing experimentation allow voluntary sub-
mission to experimentation by non-prisoners but exclude all prisoners on
grounds that prisons are coercive environments which preclude voluntary
consent. Such a law would be based on a false belief which caused it
to discriminate arbitrarily against class (1) prisoners. The law has
mistakenly used the irrelevant property of being a prisoner rather than
the relevant one of being in an inherently coercive environment. More-
over, the mistake would be compounded if the members of other inherently
6-15
coercive environments (e.g. military training camps, asylums for the
mentally infirm, etc.) were not similarly excluded. In general, rules
and laws are unjust when they make distinctions between classes that
are actually similar in what is actually a relevant respect, and/or
fail to make distinctions between classes that are actually different
in what is actually a relevant respect. This is a simple but important
policy application of the formal principle of justice.
IV. The Principle of Fair Opportunity
Material principles of distributive justice which state relevant
properties of proper distribution have until this point occupied the
bulk of our attention. Consider now those properties which might and
often do serve as bases of distribution but which in virtually all con-
texts are not relevant to distribution, or at least should not as a
matter of justice be considered relevant grounds. I am thinking of such
characteristics as sex, race, religion, I.Q., and social status. We
do in some limited contexts use such properties as relevant — e.g., if
a script calls for an actor in a male role, then females are excluded.
Still, we wo not allow distribution of economic considerations solely
or even primarily on the basis of such properties. (Acting ability is
the primary property in the actor example.)
It is important to ask why we do not accept principles such as
"to each according to sex" or "to each according to I.Q.," as valid
principles of distributive justice. The most widely accepted reason
why we exclude such properties, and indeed regard them as discriminatory,
is because to use such principles would be "to treat people differently
6-16
in ways that profoundly affect their lives because of differences for
3
which they have no responsibility ." This reason for excluding the use
of distributive principles in some contexts has important implications,
for this understanding of justice effectively means that differences
between persons can fairly be made relevant differences only if those
persons can be held responsible for these differences .
To see the importance of this point, consider the distribution
of benefits in a state. The fair opportunity principle, as I shall call
it, says that none should be granted benefits on the basis of their (let
us say "advantageous") properties when they are not responsible for
those properties, and it also says that none should be denied benefits
on the basis of their (let us say "disadvantageous") properties when
they are not responsible for those properties. Such properties are
never grounds for morally acceptable discrimination between persons
based on justice because they are not the sorts of properties that one
has a fair chance to acquire or overcome . Of course in many societies
properties such as religion and social status may not be acquired, and
can be "overcome." I am thinking primarily of race, sex, and I.Q. —
those properties that seem to bedevil fair treatment more than any others
known to the human race. But however that may be, the fair opportunity
principle is probably best treated as a second-level principle of dis-
tributive justice.
3
W. K. Frankena, "Some Beliefs about Justice," The Lindley Lecture , as
quoted in Joel Feinberg, Social Philosophy (Englewood Cliffs, N.J.:
Prentice-Hall, 1973), p. 108 (italics adde'd) .
6-17
Consider in this connection how the fair opportunity principle
could be used in direct application to problems of the institutionalized
mentally infirm. If IQ is something for which a person is indeed not
responsible, and if none should be denied the benefits of the state (or
other distributional system) on the basis of any such property, then it
would be unjust not to distribute to them the benefits generally conferred
upon all who share in the system of benefits. But this is still some-
what vague and in need of theoretical amplification. Consider the specific
example of a basic education. We confer this benefit on all citizens
equally, and we would consider a person deprived or harmed if he were
unable to receive it. Suppose that in a given community an efficient
school system is maintained which gives a uniform opportunity for a high
quality education to all, regardless of sex, race, religion, etc. Such a
system, however, will not confer an education upon those students with
reading difficulties or mental deficiencies. They require special train-
ing in order to overcome their problems and to receive what for them is
a minimally adequate education. If such persons were responsible for
their slowness, we might say that they deserve no special training and
simply have to expend more effort. But it is precisely when we discover
that they are not responsible that we say they deserve special consideration;
and hence we introduce different kinds of education for different kinds of
students regardless of the differential in cost (within limits, of course).
Notice that it is not the economic consideration of an equal dis-
tribution of resources which is being proposed. We wo not say that IMI's
(or slow learners with special reading problems, etc.) should get the
same amount of money or training or resources as other pupils because they
6-18
are not responsible for their condition. Rather, we say they should
get what for them is a quality education, even if it costs more, be-
cause the principle of fair opportunity requires that they receive it.
Any other system of distribution would lead to the consequence that an
undue burden is being placed on this class of persons. The burden is
undue because placed in violation of a principle of justice.
The argument just presented is a sketch of a justification of un-
equal distribution to the institutionalized mentally infirm which is
based on the (second-level) principle of justice which I have called
fair opportunity. The argument has been that you cannot introduce
criteria such as effort and merit in cases where a person is handicapped
through no fault of his own. Determining a person's due or desert
exclusively on the basis of such principles would in fact be morally
wrong; for, as applied to the handicapped, such principles do not
satisfy the requirements of the principle of fair opportunity. It is
not my intent to actually defend an unequal distribution of resources
to IMI's. I have used this argument merely as an example of the im-
plications which principles of justice have for problems of public
policy.
6-19
This paper discusses various problems of distributive justice,
with emphasis on the major distinctions, principles, and methods of
moral argument which are shared by most who have written on the
subject. No attempt is made to develop the "right" theory of dis-
tributive justice, but a definite framework for understanding the
wider implications of any particular theory of justice is provided.
The paper deals in order with the following topics:
I. The Concepts of Justice and Distributive Justice
II. Principles and Theories of Distributive Justice
III. "Relevant" Respects and Classes of Subjects
IV. The Principle of Fair Opportunity
6-20
7
THE IDENTIFICATION OF ETHICAL PRINCIPLES
James Childress, B.D., Ph.D.
The Identification of Ethical Principles
James F. Childress
Kennedy Institute
Center for Iiioethics
Congress has charged the National Commission for t he Protection of Human
Subjects of Biomedical and Behavioral Research "to conduct a comprehensive in-
vestigation and study lo identify I lie basic ethical principles which should
underlie the conduct of biomedical and behavioral research involving human
subjects." This paper focuses on one part of that charge — the identification
of ethical principles — and analyzes some of the issues raised by attempts to
classify some considerations, reasons, arguments, etc, as "ethical" or "moral"
and in distinguishing them from others such as political, legal, and theological.
While indicating some of the major options regarding tltese issues in contemporary
ethics, I shall develop my analysis with special attention to policy-making and
draw examples from that activity,
I. Setting of the Question
As a preliminary matter, it is important to inquire into the presuppositions
of the question "how to identify ethical considerations or action-guides," and to
1
suggest exactly how I shall approach it and attempt to avoid certain pitfalls.
Perhaps Congress attached little or no significance to "ethical" as a modifier of
"principles," as it may have been interested in what ought to be done without
worrying too much about the sorts of reasons that could be invoked. On the other
hand, assuming that Congress attached some importance to "ethical," one should
try to determine what sorts of principles would be thereby excluded from considerations
and how "ethical" ones are distinguished from and related to them,
But why would one be interested in distinguishing sorts of considerations
instead of dealing with the substantive arguments for policies however they might
be classified? Why not ask whether' there are good reasons for a pol icy without-
worrying about whether those reasons can be classified as ethical, or legal, or
7-1
political, or theological? Interest in the question presupposes a certain
historical and social context marked by what Talcott Parsons and others have
2
called "differentiation." The western world in particular has been altered
by a process of differentiation which rendered activities and institutions
distinct and functionally specific. Thus, one can see how politics, law,
economics, and religion, for examples, have become different tatt-ed from eacli
other; they no longer reside comfortably under a "sacred canopy. 11 Functional
rationality came to legitimate the norms in various institutional areas such
as politics and economics. This process of differentiation was, also a process
of secularization in one sense as "autonomous institutional ' ideologies ' re-
placed, within their own domain, an overarching and transcendent universe of
3
norms." In another sort of world, with less pronounced differentiation, questions
about the classification of considerations for policies would appear to be less
important. In such a world agreement about the releva nce of certain considerations
(such as religious ones) for policies could be assumed even when there was dispute
about the substance.
Our interest in the classification of reasons may we] 1 reflect a belief that
certain considerations are not relevant to public policy. A3 though we 'are in-
terested in the terms of public debate in a pluralistic society, we must be
wary of supposing that classifications of reasons can finally prove conclusive in
substantive debate — a supposition that may motivate some of the interest in this
question. To distinguish "moral" or "ethical" considerations from other sorts
(which we will label "nonmoral") is not to determine how much weight those sorts
of reasons should have unless one writes certain inappropriate evaluations into
the definitions of "moral" or "ethical." Thus, we shall be wary of "ideological"
definitions that write a substantive moral conclusion into the distinction between
moral and nonmoral considerations. The distinction between "moral" and "nonmoral"
should not be confused with the JJ.stlncl Ion between "moral" ^pd "Immoral," The
former is a "second-order" distinction, and the task of analyzing it is often
labelled metaethics in contrast to normative ethics. Although the distinction
between metaethics and normative ethics is controversial (and some suggest
that some metaethical views imply normative ethical ones), I am interested in
what conditions must be met for a position or argument to be considered moral
rather than nonmoral. I won't deal with the criteria of "good" and "bad" moral
reasons.
Attempts to distinguish moral and nonmoral reasons may be primarily re-
portive or reformative, A reportive definition describes and elucidates contem-
porary usage, while a reformative approach proposes a particular way of distinguishing
"moral" and "nonmoral" even if it modifies ordinary usage. Working descriptively,
a philosopher might ask what our ordinary concept of morality is or entails,
"what we actually mean by 'moral' and 'morality' in their relevant uses, or what
5
the prevailing rules are for the use of these terms." The reformative approach
would also draw on contemporary usage, but it would not be bound by it.
Most philosophical discussions of "morality" appear to be reformative or
revisionist in intention, even when they are not so acknowledged, and I too will
be most interested in recommendations about how we should use the terms. People
make such recommendations in specific contexts and for definite purposes. Why
do we want to distinguish moral and nonmoral action-guides? A moral philosopher
might be interested in expanding or contracting the scope of Ills legitimate en-
deavors; for instance, a moral philosopher may think that it is appropriate to
discuss "personal ideals" and he may argue that a definition of morality must be
6
broad enough to encompass them. What such a philosopher considers in defining
morality is certainly relevant to the policy-maker who is also interested in
the classification of arguments. 'But why one wants or needs a definition of
morality is not irrelevant to determining the sort of definition that ie
appropri ate.
7-3
Even for a revisionary definition of morality, proposed for some fairly
specific purposes and judged by its usefulness particularly in illuminating
certain issues, ordinary usage remains an important but not decisive test.
The proposed rules for using the terms should bear some close relation to
ordinary usage, or substitution of terms would seem wiser. But contemporary
usage is diverse and perhaps even imprecise and inconsistent. Joel Felnberg
comments that "[t]he word 'moral,' reflecting a variety of disparate and con-
trary uses in the technical literature of law, theology, and philosophy, is not
simply ambiguous, but ambiguous in such an extraordinary way that some of its
senses are antonyms." His point holds for ordinary usage and not merely for
technical literature. He also gives several examples,
On the one hand. .. 'moral ' has the ring of supreme authority,
and on the other, it still carries its original sense of
informal 'customs' or 'ways.' Aristotle contra s ted the moral
with the intellectual; Kant identified it with the ration al .
Lawyers use it to refer, on the one hand, to loose and in-
formal arrangements beneath the official attention of the
law and, on the other hand, to an ultimate standard for
appraising the law. 7
Furthermore, as I shall emphasize later, "moral" sometimes refers to decisions
and acts that should be left to individuals or small groups, because they are
not sufficiently important to warrant the community's interference, and some-
times it refers to the issues that are considered so important to the community
that it expresses certain judgments through laws and policies. This particular
ambiguity is closely connected with the main options in the debate about the
formal and material criteria of "moral" over against "nonmoral."
These ambiguities in contemporary technical and ordinary usage suggest
that attempts to delineate the necessary and sufficient conditions may be
doomed to failure. At the end of a search for those conditions for Identifying
reasons, considerations, and action-guides as "moral" in contrast to "nonmoral,"
7-4
one may simply have to conclude that "moral" and "morality" are "family-
resemblance" words (Wittgenstein's phrase). They may be so open-textured
that necessary and sufficient conditions cannot be identified. Nevertheless,
an examination of some recent philosophical efforts to identify necessary and
sufficient conditions of "morality" will be instructive.
11. Formal ami Material Conditions of Mor.illiy
W. D. Hudson, a philosopher, offers the Parliamentary debate on the Divorce
Reform Bill in 1969 as one example of first-order moral discourse in which policy-
makers try to determine what they ought to do (we could easily recall or imagine
others regarding abortion and euthanasia as well as research involving human
subjects) :
This bill recognized the breakdown of marriage as the
sole ground for divorce. .. .The moral question which
arises is: ought divorce to be reformed along these
lines? That was the question debated at the all-night
sitting. .. .We may safely take this debate as an example
of moral discourse. The participants put forward what
purported to be their omanswers to a moral question.
They supported these with what they considered reasons
appropriate to the settlement of a moral issue.
Hudson makes numerous assumptions about this debate in terms of its characteri-
zation as moral, and they may well be warranted, but whether other interpreters
would also view the debate as involving "moral question" and as "an example of
moral discourse" depends on their conceptions of "moral" and "morality." To
ask "what ought we to do?", "what law or policy ought we to have?", or to judge
"that proposed policy is wrong," may be to ask a moral question or to make a
moral judgment. Clearly the judgment, for instance, is normative , but there may be
many normative judgments and action-guides (such as etiquette) that are nonmoral.
Our question is not which morality, counts and how much it counts but what counts
as morality.
Recent philosophers have held that one or more of several criteria determine
what will count as a morality or moral act ion-guide. The main criteria arc that
7-5
an action-guide must be (1) prescriptive, (2) universalizable, (3) overriding,
and (4) indicative of a material content, particularly one that relates to
9
effects on others. Few philosophers quarrel with some notion of prcscriptivity ,
except as it is filtered through the other formal features of universal inability
and overridingness , and I shall thus emphasize 2 and 3. 3-3 are formal criteria
that are taken to apply to any moral action-guide regardless of its content.
Not what is held, but how it is held determines whether it is moral or not. 4, on
the other hand, requires certain material content before the action-guide can
qualify as moral. The formal criteria, taken singly or jointly, will, of course,
admit a broader or wider range of judgments, as far as content is concerned,
whereas #4 recognizes as moral only those action-guides, etc., that take account
10
of some specified content. Particular philosophers may, of course, claim that
one or more of these conditions are necessary but not sufficient, necessary and
sufficient, or sufficient but not necessary for applying the term "moral" to
action-guide.
Consider the following arguments between government X and opponents of a
war that X is undertaking, Government X claims that the war is just and that a
policy of conscription is also just; its claims meet both formal and material
criteria of morality. Among the opponents and resistors are
A contends that the war is wrong (but his judgment
satisfies only formal criteria of "morality") .
B contends thafthe war is wrong (but his judgment
satisfies only the social material criterion of
"moral ity") .
C contends that the war is wrong (and his judgment
satisfies both t he formal and social material
criteria of "morality").
D refuses to serve because military life with its
hierarchy and its emphasis on obedience is incom-
patible with his ideal of human excellence.
E doesn't want to risk being killed,
7-6
F insists tliat the war is not in the national interest
and refuses to serve in it.
F condemns the war because the evil results will outweigh
2
the good ones.
F argues that the war can't be won.
F claims that the war lias not been authorized by the
proper political authorities (cf. CI).
G holds that the war is illegal and unconstitutional
(cf. ty.
H says that only one war can be just — the battle of
Armegeddon which God will command and direct,
I choose this case" as a focal point because we have all heard (and perhaps made)
many of the above arguments in the last several years and because we probably
also have some rough idea how we would classify them — moral, prudential, political,
legal, religious, etc.
With this case in mind, let us turn to a more detailed discussion of formal
and material conditions of morality. Among the formal conditions I shall em-
phasize universalizability and supremacy or "overridingness . " Hare and many
others construe universalizability as a logical requirement, not a substantive
moral one: to call a judgment universalizable means "only that it logically
commits the speaker to making a similar judgement about anything which is either
exactly like the subject of the original judgment or like it in the relevant
respects. The relevant respects are those which formed the grounds of the
original judgements." We cannot make different m oral judgments about acts
that we consider to be exactly or relevantly similar. One might base this
thesis on the logic of "moral" or "oughts" or judgments generally (including
both evaluative or normative and descriptive judgments). Universalizability
does not distinguish mora_l_ Judgments from other sorts although it may be a
necessary condition for any normative judgment, Hare apparently thinks that
universalizability together with prescri ;>t ivi ty and perhaps supremacy are
sufficient as well as necessary for mor a I judgments and action-guides.
7-7
The formal approach has been criticized on several grounds. First, it
casts its net so widely that it catches numerous action-guides that we ordinarily
consider to be nonmoral. In particular, the criterion of supremacy, overr idingness ,
or superordination lias a paradoxical effect. Hare contends that in one sense of
"mornl" ("perhaps the most important one"), a "man's moral principles. , .are those
which, in the end, he accepts to guide his life by, even if this involves breaches
12
of subordinate principles such as those of aesthethics or etiquette." Although
Hare appears to be working with a fairly clear distinction between moral and af
least some nonmoral value judgments (such as aesthetics), both share universal-
izability and prescriptivity and so could only be distinguished by the supremacy
of the moral judgment. It is difficult to see how the formalist can avoid tine
odd consequence that if someone holds aesthetic judgments as supreme, as over-
riding all other judgments, he must have taken aesthetic judgments as his moral
ones. Whatever action-guides a person takes to be supremely authoritative or
as his way of life would appear to constitute his morality. For instance, a
biomedical researcher whose life is devoted to scientific progress could con-
ceivably be said to have scientific progress as his morality. While there is
some warrant fci sucli a view in ordinary usage, it obfuscates some of the usual
distinctions between religion and morality and, perhaps, would permit even a
thoroughgoing, serious prudentialism to be considered moral (as contrasted to
nonmoral) if its proponents took it as supreme and overriding and if it met the
other formal conditions of morality.
The criterion of supremacy of "overridingness" at least in its absolute
sense appears to gloss over a significant human problem (to call it a "moral"
problem at this point would beg questions) in that it prejudges the dominance
of particular sorts of action-guides or reasons. Much of our serious debate
concerns which so rts of considerations should be given the most weight, for
7-c
instance, prudential or moral ones. Thus the debate about whether a position
is moral is simultaneously a debate about how much weight and significance it
should have, but the formalist won't let the content determine whether the
position is moral. As Hare says, "there is absolutely no content for a moral
13
prescription that is ruled out by logic or by the definition of terms."
This feature of "overridingncss" (or other terms such as ultimacy, supreme
authoritativeness, and finality) appears to rule out another fundamental human
question: "why be moral?" For if, by definition, a morality or moral
action-guide is supremely authoritative, it makes no sense to ask "why should
14
I/we abide by it?"
The use of the formal criteria, particularly universalizability , has been
criticized as hopelessly "culture-bound," "ethnocentric," and even "excessively
Protestant" and "individualistic." Most of the charges hold that universalizability
cannot be construed merely as a "logical" thesis, for it actually builds in a
particular morality. It pretends to be neutral, but it actually chooses sides
within morality, hiding its allegiance to "liberal morality" under the guise of
15
neutral, logical analysis. The requirement of universalizability fails to take
account of radical moral disagreement (e.g. whether to take the standpoint of
impersonality). It sets the terms of moral debate and the weight of sorts of
considerations without taking responsibility for doing so. Although I share
this concern of many critics of the universalizability requirement, I am not
convinced that the requirement must be surrendered. It is clearly not sufficient,
but it does appear to be a necessary condition of moral judgments. It begs
fewer questions if we clearly separate it from the condition of supremacy.
One advantage that could be claimed for the formalist understanding of
morality is that a number of social conflicts whose participants offer pre-
scriptive, universalizable, and overriding judgments can be construed as mor;*l ,
Thus, Hare describes a debate between a Mazi and a liberal as a moral conflict.
7-9
Neverthel ess , it docs not seem to be necessary to concede that both (or all)
i
sides in a debate are offering moral considerations (as contrasted to nonmoral
ones) In order to view the debate as a moral one; this point would seem to
apply to the Parliamentary debate about the Divorce Reform Bill that 1 mentioned
at the beginning of this section. A debate may be classified as mora] because
it is, in part, a debate about whet her to be moral and to take the perspective of
morality, or because it is a debate between a recognizable moral position and a
nonmoral one (such as self-interest), as we do not hesitate to consider the
16
personal conflict between duty and interest a moral one. When we classify as
moral a debate between government X which claims that it is engaged in a just
war and its opponents who hold that the war is wrong, we do not necessarily
ascribe a "moral" claim to both parties.
The formal criteria of morality, as we have seen, appear to allow us to
count as moral a wide range of judgments and action-guides, including some
that would ordinarily be thought of as nonmoral. Several philosophers have
proposed a narrower concept of morality that has some support in ordinary speech.
It is both social and mate rial ■ It focuses on regard or consideration for others,
particularlv for their welfare. Of course, some of the formal conditions such as
universalizability involve some sociality (in so far as they insist on treating
similar cases in a similar way or legislating for others) , but the narrower con-
cept of morality writes in "other-regard ingness" or consideration of others.
Furthermore, this narrower concept is m ateria l as well as social as it builds in
regard for the e ffects on others' welfare . It characterizes morality by its
"subject-matter," "its content, what it is about, the range or type or coan-
17
siderations on which it is founded." From this perspective, formal conditions
may be necessary but they cannot be sufficient; a moral judgment or action-guide
must also meet a material condition.
7-10
As in the discussion of the formal conditions I will only be able to deal
with a few salient points about the social material condition. Furthermore ,
particular versions of this condition (just as particular versions of the formal
conditions) may escape some of the objections which I try to state in general
terms.
One major difference among the versions is t ho degree oT specificity of
effects on others' welfare. If one can be fairly specific about those
effects' and about welfare, obviously one can draw sharper boundaries between
"moral" and "nonmoral." G, J, Harnock contends that the notion of "'the welfare 1
of human beings surely has... a perfectly clear and determinate core of centre,"
even while it also has "an extensive penumbral fringe of vagueness and indeterm-
nacy." One can say that torture, starvation and humiliation are bad as a matter
of fact, not opinion; likewise, one can say that love and care are better than
18
hatred or neglect. Rejecting such criteria as happiness, others identify welfare
19
interests as those associated with survival, security, health, etc.
Although he refuses to offer any specific content to the effects on others
that must be considered, Frankena adds a clause that requires explication. He
says that for an action-guide to be moral it must include judgments, rules, etc.
that "concern the relations of one individual. .. to others" and "involve or call
for a consideration of the effects of his actions on others (not necessarily all
others), not from the point of view of his own interests or aesthetic enjoyments,
20
but from the 1 r own point o^f view . " Functionally, this requirement of consider-
ation from the other's point of view is similar to the notion of a determinate
core of welfare, which it is presumed the other would accept. Frankena ' s re-
quirement rules out egoism or prudentia 1 ism if these are understood as basic
perspectives (instead of being derived from a conviction that the welfare of
Others could best be protected by each person's pursuit of his own interests).
Egoism or prudent in 1 ism could be a "workinn criterion" of morality but not its
7-11
basic principle. This social material condition, as sketched by Frankena,
"rules out as non-moral only such AGs as pure egoism or prudent ialism, pure
21
aestheticism, and pure religion."
A number of objections to making the social material condition necessary
hold that it builds a substantive moral position into the definition of morality.
The charge that this criterion builds in "utilitarianism", for example, may be
more devastating for some specific statements of the content of human welfare
than for more general statements, which leave open what effects on others must
22
be considered and whether deontological considerations are germane.
The social material condition, at least on some interpretations, does not
rule out particularistic moralities, for it does not require that effects on
all others be considered, or that all others be counted as humans, but only that
the effects on the welfare of at least some others be considered. Thus, it does
not exclude the Nazi from the moral camp, for an argument on behalf of acts and
policies that inflict suffering on some for the sake of others still qualifies
as "moral" (in contrast to "nonmoral") although we may insist that the acts and
policies are morally unjustified. Thus, if someone recommends a policy of massive,
high-risk, non-therapeutic experimentation on prisoners on behalf of future generations,
however much we may morally condemn that recommendation, we cannot dismiss it as
nonmoral on the social material condition of morality. Indeed, it refers to a
determinate core of welfare (the health of future generations) and it includes
the effects on some others from their own point of view (what future generations
need — not merely want) . Considerable moral disagreement remains because of the
necessary balancing or weighing of the welfare of one group against risks to another,
the welfare of present vs. statistical lives, immediate vs. remote harms and goods, etc.
Although I am more interested in highlighting several issues than in de-
fending ii particular interpretation", I think that a strong case can bo made
that formal conditions of prescr iptivity and universal izabf 1 ity and the social
7-12
material condition of "other-regardingncss" are necessary and sufficient for
moral judgments and action-guides. Because it is necessary to leave open for ,
debate whether we should abide by moral or nonmornl action-guides, particularly
when they come into conflict, I would omit the formal condition of finality,
23
supreme importance, or ultimate authoritativcness . Only by emphasizing that
human discourse remains open, even if the term "moral" is restricted by the
above criteria, can one partially blunt the charge Uihat one is making logical
or conceptual analysis do the work of normative ethics or choosing sides in a
moral debate while claiming to be neutral.
III. Ideals
Reslster D presents his reasons by depicting his ideals . Although some
ideals (for example, charity) may be otherregarding, D's ideal is not. It is
an ideal of human excellence that makes no refeience to the welfare of others;
he simply doesn't want to be a subordinate in a military organization although he
is not upset about killing in war or fighting in a putatively unjust war. His
argument is similar to aesthetic ones and he might claim that participation in
a military organization is "degrading." According to the social material criterion,
his ideal is nonmoral, while according to the formal criteria, it could be moral.
Numerous factors that are not universalizable, etc., shape personal and
institutional decisions, but I am only interested in those factors that can be
invoked as reasons for decisions. The question is wha t form and content those
reasons must have to count as moral and where ideals fall. Strawson has dis-
tinguished the regions of the "moral" and the "ethical," including social duties
among the former and ideals among the latter, but this distinction which has
little or no support in ordinary usage is unhelpful. From the standpoint of moral
discourse, what are we to do with 'Ideals? Do ideals constitute a level with in
morality or beyond it? What difference does it make whether we characterize them
24
as moral or nonmoral?
7-13
Hare would consider D's reasons for his net as moral largely on the grounds _
that people do not hesitate to call sucli reasons moral, that moral philosophy ;
should deal with them, and that ideals may be universalizabl e, prescriptive, and
overriding. Their role in conduct, judgments, and justification is analogous to
the role of moral principles. Others might say that ideals (as well as other
"personal oughts") arc at least similar to and akin to moral oughts. Although
Frankena, on the other hand, is inclined to view non-otherregard ing ideals as
nonmoral, he thinks that they can be a legitimate subject of moral philosophy
which must consider, for instance, whether such ideals have priority over moral
principles in cases of conflict. Here again because the discussion in philosophical
literature is partly concerned with the appropriate scope and subject matter of
moral philosophy, some of the points may not be important to the policy-maker
25
who is interested in the classification of arguments for somewhat different reasons.
One option is tempting but ultimately unsatisfactory. One might say, for
instance, that a debate between observers (perhaps a (clemency board) about whether
D acted wrongly, should be pardoned, etc. would be moral because D's act has direct
26
and indirect effects on others whether he recognizes and considers them. Thus,
these observers would be making moral judgments if they considered these effects
on others or if they decided how much weight to give to government X's moral
judgments (assuming their validity) in relation to D's nonmoral judgments. These
observers could well be engaged in moral debate over a moral issue. But such a
move would not deal with the question from the agent's standpoint — whether D is making
a moral argument whatever he calls it. What is at issue is whether D supposes that
the pursuit of his ideal relates to human welfare. If we assume that the ideal
does not involve otherregardingness and thus is purely personal, the argument be-
tween D and X could still be considered moral, for it concerns whether to be moral
and whether X's moral reasons should take precedence over D's personal ideals.
One last point about ideals before lurning to prudential and political judg-
ments. .Social ideals are often invoked in political debate. Most often they
7-14
have some reference not only to society (by definition) but also to effects on
members. Nevertheless, there may be a social ethos that stresses the universal
participation of all its citizens in military or alternative service or non-
therapeutic research for quasi-aesthetic rather than moral reasons. The community
might ask "what ought we to do?"but answer it with an answer to another question
"who are we?" It will not do, I think, for an observer to say "that's obviously
a moral issue, since following that ideal has effects on the welfare of others,"
For the representatives of the society may see it as a matter of ideals apart from
those effects, and may be concerned with realizing those ideals or with expressing
them in policies. Particularly in the language of expressing ideals, one finds
an emphasis on policies that symbolize the community's ethos. For instance, in
the recent debate over amnesty, many policy-makers and citizens asked what parti-
cular policies would symbolize about the nation: that it is vindictive or respectful
of law and order, that it is forgiving or soft? Although the same policy may be
justified by the same people in terms of a country's ideals and its moral standards
(including the social material condition), and although both reasons usually
appear together and are often hopelessly entertwined, the language of "public
ideals" needs more attention. Such ideals bear some close kinship to morality,
even when they don't meet the social materia] criterion, and they frequently
28
present at least "semimoral" oughts.
IV, Prudence
Objector E opposes and refuses to serve in the war because of his prudential
calculation of risks to himself. He may take his egoism as overriding, prescriptive,
and universalizable, and, if so, his position is "moral" according to formal eri-
29
teria, at least according to some interpretations. On the other hand, he may not be
basically egoistic, but rather derive hi s working egoism from a more basic principle
that the general human welfare would be served by egoistic actions or that the world
would be better off if everyone acted as he did. The latter version of his position
7-15
would qualify as moral on some Interpretations of the social material con-
dition as well. Furthermore, even Frankena includes some duties to oneself
under morality although it is not clear whether he claims that a system of
morality must have some action-guides that are at least in part otherregarding
or that e ach action-guide, to be m oral , must be at least in part otherregarding.
But suppose that E is basically prudential and egoistic — he only wants to
save his 'own skin. The issue between government X and E could still be con-
strued as moral in the sense of being a debate about whether to be moral and
about what weight morality should have on the scale of reasons for action. But
E's own position would not be considered moral, for, as those who have insisted
on the condition of otherregardingness contend,
'the moral point of view' involves, precisely and essentially,
the ab andonment of pure prudential egoism, and a readiness to
consider as justifying grounds of action the interests or 'wants',
ideally of everyone, but at any rate of at least some persons
other than oneself.
A few more comments about prudence are in order. Traditionally (for instance,
in Greek thought and St. Thomas Aquinas), "prudence" lacked the negative connotations
it has since acquired. It had a different ring: "the pre-eminence of prudence...
f inally. .. signifies the directing of volition and action toward objective reality.
The good is prudent beforehand; but that is prudent which is in keeping with
reality ... .The immediate criterion for concrete ethical action is solely the im-
32
perative of prudence in the person who has the decision to make." The radical
change in the value of prudence resulted from the radical alteration in the under-
standing of reality, of the self and its world. For Machiavelli (The Prince , XXI),
"(p)rudence consists in being able to assess the nature of a particular threat and
33
in accepting Lhe lesser evil."
There are many precepts of self-protection, of which prudence is only one.
If a person refuses to participate In a perilous experiment, he may be acting
7-16
prudently In the sense of taking few or no risks. "To act prudently is to
play safe, for near-certain gains at small risks." But the precept of not
wasting one's abilities is sterner than the precept of prudence. Furthermore,
the prudent cannot be equated with the e xpedient , It may be prudent to save,
but it may not be expedient to begin now; "reasons of expedience are reasons of
a special sort: reasons for doing something on the ground that it Is incidentally
at hand to serve one's purpose, or because it serves a purpose quite incidental
34
to the purpose for which we would normally be doing this thing,"
Falk contends that "personal oughts" such as prudence "all seem at least
akin to a 'moral' ought in their action-guiding force and function," but lie is
hesitant to speak of them as more than "semimoral." "Oughts" that concern the
interests of others seem to have a more stringent obligatoriness, partly because
of their end and partly because, as Kant emphasized, an ought predicated on the
assumption that one is pursuing an end (such as one's health) is at best hypo-
thetical: "if (or as) you want to be healthy, you ought to get enough sleep, eat
the right foods, etc."
Falk's main contention is sound; if we define "moral" to include a social
material element, we might by formal criteria still consider some personal oughts
such as prudence "akin" to "moral" or even as "semimoral." However we define
"moral," we should not conclude that "personal oughts" are always inferior and
subordinate to "moral oughts." That should not be a matter of definition. I
think that little is gained by calling basically egoistic positions moral in-
stead of nonmoral, as long as we recognize that this classification does not
determine their weight and leave open for debate the :;ort of perspective we
ought to adopt.
Furthermore, there is a form of "social prudence" that involves calculations
about the good for a larger group sucli as a political community. One meaning of
"politic" is "prudent," and one may deUimine the "politic" or "prudent" course
7-17
of action, especially in adapting means to ends. How does this form of social
prudence relate to "moral"? Although my discussion can only be sketchy at best.,
I will consider this point in relation to "political" judgments, reasons, and
considerations.
V, Politics and Political Considerations
Among the possible positions in opposition to or refusal of service in a
particular war, F - F, could be considered "political" in nature. We might say
of F ("the war is not in the national interest") that it is "purely political,"
claiming that it does not refer to any considerations but political ones. Although
ordinary usage would appear to support such a statement, neither the formal nor the
social material criteria of morality require that F be classified as "purely
political" in the sense of totally "nonmoral." One difficulty, of course, is
transposing criteria that might be clear for "moral" judgments and considerations
about individual conduct in interpersonal settings to the political order. Never-
theless, policy-makers and citizens could construe the national interest as a
"moral" consideration (or at least a "semimoral" one) since it can be universalized
(on some interpretations), prescriptive, and may be overriding-. Furthermore,
national interest may not violate the standard of otherregardingness, for it in-
vokes the interests of some persons other than the agent. The same point could
perhaps be made about group interests (e.g. labor unions) in pluralistic politics.
But such interpretations are somewhat strained.
One possible argument for classifying F 's judgment (and F 's too) as political
is that (1) it is based on the sorts of considerat i cr.s that policy-makers examine
and the type of reasoning they employ, and (2) it is the sort of matter that can
and should be resolved through the , p ol it ic al proc ess.
l.ct me start with the second reason, which requires several comments. First,
as Giovanni Sartorl contends, "in order to find our v.v.y in the differentiations
among politics, ethics, law, economi a mui so forth, ir is necessary to refer to
7-18
the structural differentiation of human aggregations. Perhaps because of a
lack of categories, perhaps for other reasons, the fact is that only the field
of ethics, which is the most ancient and the most developed, escapes reference
35
to a structural underpinning." In contrast to "ethical," the other terms such
as "political and "legal" signify a locus, a site of behavior, a system, which
is "a constellation of structures, roles, and insl i tut Jons. " Thus, 1l is necessary
to refer to that sytem in discussing "political" judgments, in distinction from
"ethical" ones, but this necessity poses a special danger: it is difficult to
state the criteria of "political" without implying a specific type of political
system such as a liberal, constitutional democracy. If "political" is understood
as referring to a system within the social structure, F - F but also most of
1 4
the other positions could be understood as "political" in that they fall within
the political sphere, are directed toward particular policies and laws which they
want to change, etc. But this sense of "political" would not deprive any of these
positions of the label "moral" or necessitate their classification as "nonmoral."
Indeed, they could be both moral and political.
Second, we establish procedures and processes for reaching certain decisions.
Admitting that "in a sense all decisions are mora] decisions," Virginia Held goes
on to emphasize that "we seem to make moral decisions to treat given ranges of
decisions some other way, to adopt more limited decision procedures for sucli
36
ranges." We assign a number of matters to the political process, to be determined
through established procedures (such as voting, etc.), and, within limits, abide
by the outcome without reexamining the procedures and process each time. On these
grounds F. might properly be viewed as political. Such a judgment can only be
made within a particular political system (or within a proposal about a political
system). Sometimes, however, we may feel that constitutional or moral limits (cf,
CI and B as well as a version of C that insists that the war is being conducted
unjustly) have been violated, and in soii-.i cases agents may make moral and political
7-19
("moral-political"?) decisions to oppose those outcomes and perhaps even to
work for a new system, (The dispute about wh i c h political procedures are de-
sirable is itself a political dispute.)
Certain types of consid erations seem to belong, to political procedures and
processes and claim the attention of political agents, who usually reason in
certain ways. If there were no disagreements about ends, goals, and purposes in
human life, there would be no politics, Among angels and beasts there is no
politics — among angels, presumably, there is complete concord of purpose; among
beasrithere is no capacity for projecting ends, The context of politics and
political thought is the collision of some human ends and purposes. Where such
a collision is absent, the only remaining questions are technical; how do we
achieve our goals? But such questions are for the experts. The political
question par excellence is "what ought we (not I) to ccJo"?
Variously viewed as a problem of order and freedom, unity and diversity,
public and private, political life is a matter of "continually creating unity,
a public, in the context of diversity, rival claims, unequal power, and conflicting
interests." Although the enterprise of politics is conducted within certain pro-
cedures and by certain processes, it is aimed at "the conciliation of differing
37
interests," Its primary mode of reasoning and justification is te leological
(right and wrong depend on the ends and consequences of actions and policies) ,
Its reasons rest on ends and consequences. But only a doctrinaire view of politics
would insist that deontological considerations (that some things are right or
wrong independent of their ends and consequences) are totally excluded, Deonto-
logical considerations (whether moral, legal-, or constitutional) may well set limits
on the ways certain ends are pursued; among the principles of a political order may
be certain convictions that set limits on what the community can do in' achieving
other goals. Tor instance, the right of free speech is not to be abridged. Such
deontological considerations may he JJjrlnw facie rather than absolute requirements.
7-20
Even if telcological reasoning is primary in political discourse and argument,
it is not a necessary or a sufficient condition of "political." For moral reasoning
apart from politics may well be teleological , and political reasoning may well in-
clude deontological considerations,
F which argues that the war is wrong because it can't be carried through Lo
victory appears to be making a Judgment of political prudence. His pragmatic
calculations are of the sort that we may (morally) decide should be resolved through
the political process since people of good will may have different assessments of
probable success. Perhaps F is reducible to F on the grounds that it is not in
the national interest to risk political survival, independence, etc. for a losing
cause. It may be close to F in that it is a judgment of proportionality although
its cost-benefit analysis is limited to one's own nation. In either case, its
classification would probably follow the pattern of F ,
F which insists that the war is wrong because the evil results will outweigh
the good ones is an example of consequent ialist reasoning. What are the grounds for
calling it "political" (in the sense of "nonmoral")? In contrast to F, which argues
only from the standpoint of the national interest, F is a prescriptive, overriding,
and universalizable judgment about the effects of policies on others and thus
satisfies both the formal and social material conditions of morality. If we
classify F„ as "political" (as often happened during the Vietnam War), we seem to
be implying the following points. Given our commitments to democratic procedures,
the question about the proportionality of good and evil results of policies is
properly political, in the sense that we have (morally) established certain pro-
cedures and processes for arriving at such decisions. Furthermore, that question
can only be answered by a mode of calculation that is properly political. If, however,
F„ argued not a point about proportionality but rather a point about the war's con-
duct (for instance, as a violation of the principle of discrimination between
combatants and noncombatants based either on moral I ly or International law), his
7-21
position would be deontological and more properly moral, Such uses of tlie label
"political" for judgments and considerations obviously presuppose commitments to
resolving certain conflicts within political procedures or to viewing certain types
of considerations as properly political. Our discussion of the criteria of "moral"
in contrast to "nonmornl" does not provide any grounds for a neutral, logical dis-
tinction between "moral" and "political" that would permit us to classify F~ as
purely political.
When we say that the President vetoed a bill "for purely political reasons" or
"only for political reasons," we may imply that he actually thought the bill was in
the public interest and that he would have signed it if he had not been swayed by
his interest in retaining power to yield to public opinion or some group, or by
his interest in influencing Congress to act in a certain way later. Or we might
agree with the President's decision and yet criticize his reasons as "purely" or
"only" political. In such usage, we strip the term "political" of any positive
moral connotations. "One says 'only' [as in "he did it only for political motives")
because something is done for the wrong or for not quite the right reason--done
38
for one reason where there is another and nearer reason for doing it anyway."
39
The meaning of "political", as Alan Gewirth has suggested, fluctuates be-
tween moral (e.g. Aristotle, Rousseau, and Hegel) and am oral (e.g, Machiavelli)
poles, just as the term law does. The moral pole often "subsumes ethics under
politics on the ground that while every community aims at some good, (as Aristotle
insisted) ' the state, tlie political community, aims at the supreme good'," The
amoral pole understands politics as the use of deception, farce, coercion, and
even violence to achieve tlie end of power as rule and domination. Other defi-
nitions of "political" falls between these poles. One is tlie rhe tori cal — "political"
indicates the use of persuasion, influence, pressure, etc. to obtain some end; here
the emphasis falls on influencing others' acts and attitudes. Another is the
ad min istrative — "political" indicates the application of rules to the affairs of
some community. The phrase "academic politics" may suggest the administrative,
7-22
while the description of a department chairman as a "political operator" may
suggest the rhetorical.
In the contemporary context, a number of political theorists accentuate the
moral qualities of politics and define it in those terms, while many political
scientists who are more empirically oriented define politics in terms of amoral
and rhetorical themes, stressing interest, power, and influence. Neither side
can claim a decisive victory by appealing to ordinary usage. Within the family
of words relating to politics and deriving from the same root, we can discern
different emphases. The adjective "politic" which comes very close to "prudent"
and "expedient" and the noun "politics" best fit the amoral concept with its
emphasis on interest and power. Both the adjective "political" and the. noun
"policy" are more neutral, although "policy" perhaps suggests more public and
moral concerns. Indeed, several political theorists such as Wolin and Arendt use
"political" as a substantive — "the political" — and eschew "politics" except to
40
describe current realities. The proponents of a moral definition of politics
may intend to offer not a reportive but a reformative definition, for which ordinary
usage is instructive but not critical. They may justify their definition on the
grounds that it enables us to understand and direct our activities better than
alternative definitions.
Surely the most difficult issues in the relations of ethical and political
judgments derive from certain conceptions of normativ e ethics (e,g. the law of
love) as applied to normat ive understandings of politics (e.g. the realist view
of politics as dominated by self-interest and the libido domi nandi ) . As Max
Weber's classic essay "Politics as a Vocation" stressed, there are two very
different moral perspectives on political life: Geslnnun gse t hik and Verantwort-
ungsethik . The former, variously "translated as ethic; of ultimate ends, absolute
values, and intentions, is cl early deontol oglcal ; it may even hold, fiat jus t It in ,
runt caelu m, and it is obviously inimical to the prudential calculation of con-
sequences. The ethics of responsibility, by contrast, takes acccount of aver arc
7-23
human deficiencies, calculates the costs, and is willing to use "morally dubious
means or at least dangerous ones" (especially coercive and violent ones) for worthy
ends. Yet these "ideal types" for Weber are "not absolute contrasts but rather
41
supplements." To insist that one perspective or the other be defined as ex-
clusively "moral" or "political" is to overlook the point that they are both
possible perspectives on the relations of morality and politics.
Although political reasoning and argument are largely (but not necessarily and
exclusively) teleological and consequentialist , many judgments in the political
42
arena may be classified as moral as well as political. One difficulty has to
do with group interest and national interest; if they are classified as egoism
writ large, they may be considered nonmoral (at. least on most interpretations of
otherregardingness and universalizability) . If, on the other hand, they are not
basically egoistic or if they meet the conditons of universalizability and other-
regardingness, political prudence may qualify as a form of moral judgment. Although
the policy-making process is complicated by the number of parties affected by de-
cisions, the range and magnitude of effects, etc,, the reasons for policies may
be appropriately considered as moral, not merely nonmoral or political, if the
formal and material criteria of "moral" are met. While the sorts of considerations
and the mode of reasoning are not sufficient to distinguish political judgments
and arguments from moral ones, we do assign certain matters to decision-making
procedures and processes that are "political" in that they are concerned with
the conciliation of differing interests. But the logic of the terms "political"
and "moral" does not dictate that such matters (and decisions about them) are
merely or purely political.
7-24
VI. Religion and Theology
Our war resistor // H insists that "Only one war can be just — the battle
of Armegeddon which God will command and direct," Since the current war does
not meet these religious criteria, it must be condemned as unjust. II 's judg-
ment appears to be based on nonmoral, specifically religious, reasons rather
than moral ones. But what is to be gained by this sort of classification? What
are the issues in the distinction and relations betwen religion and morality or
A3
theology and ethics?
Definitions of religion are almost as controversial as definitions of morality.
Some interpreters suggest that we start looking at religion in distinction from
morality by asking about their respective functions: while morality's function
is to facilitate human cooperation, religion's function is to deal with problems
of meaning (such as suffering and death) on cognitive, emotional, and practical
levels. Although functional analysis is suggestive, it does not take us very far,
and the major dispute about the definition of religion parallels the debate about
the formal and material conditons of "morality." Must "religion" satisfy formal
criteria such as "ultimacy" or "primacy," or must it have a certain conten t such
as "sacredness" or "transcendence"? Furthermore, can either set of criteria (or
even both sets together) be viewed as necessary and/or sufficient conditions of
"religion"?
Let's look briefly at some examples (and difficulties) of these different
sorts of definitions. Paul Tillich defined religion in terms of "ultimacy" or
"ultimate concern": "Religion is the state of being grasped by an ultimate
concern, a concern which qualifies, all other concerns as preliminary and which
I. kk
itself contains the answer to the question of the meaning of our life," The
condition of "ultimacy" (whether applied to concern or to objects of concern)
seems to be neither necessary nor stiff li Lent, for some religious objects such
as "spirits" may not be "ultimate" to tlio-ir believers and some principles (such
7-25
as the principle of utility) that we would count as non-religious could satisfy
this criterion. Perhaps a religious action-guide claims p riority or supreme
authoritativeness , but this condition, as we have seen, is also proposed as one
formal condition for moral action-guides as well (although it is the one that 1
declined to emphasize on the grounds that it begged too many questions at least
in its strong form). Any action-guide that is taken with sufficient seriousness
would seem to qualify as "religious" by this definition.
Without offering complete arguments for a definition in terms of some be-
liefs or the status of some objects, I think that a definition of "religious"
in terms of sacredness is more promising. It avoids some terms such as su per-
natural that presuppose a specific religious worldview (with a distinction between
nature and supernature) and thus are ethnocentric. Yet it seems broad enough.
David Little and Sumner Twiss define a "religious statement" as "a statement
expressing acceptance of a set of beliefs, attitudes, and practices based on a
concept of sacred authority that functions to resolve the 'oncological' problems
of incerpretabilicy . " 45
Religion, as defined, may affecC Che moral life in numerous. ways . Some of
Chose who emphasize Chis poinC insist thac ic is also necessary Co challenge Che
formal cricerion of universalizabili ty in order Co acknowledge Chis influence or,
more specifically, in order Co do justice to "Christian ethics," "Jewish ethics,"
etc. But the fact that religions may offer motivations for being moral and may
affect character, decisions, etc. is no reason for eliminating the requirement
of universalizability . Wc are interested in the process of reason-giving, in the
juscif icaCion of decisions and particularly policy-recommendations. As Frankena
insisCs, one may correcCly maintain that "many aspects in the shaping of a life,
moral or religious, need not be subject to the universalizability requirement.
Only normative and evaluative judgments are." Here the issue is not between
"religious" and "moral" but racher between judgments and ocher aspects of the
7-26
moral life which need not be subject to the uui versalizab i 1 i ty requirement in
order to be moral. Even if they meet the condition of uni versalizabl lity , religious
systems may well be normative or value systems without thereby being mora l ones,
If one takes the purely formal criteria, religious action-guides that are
universal izable, prescriptive, and overriding could qualify as moral. Excluding
the criterion of universnliznbi 1 ity (or at least some of its versions), .lames
Gustafson holds that since religious action-guides are prescriptive and overriding,
they should be counted as moral or not merely as nonmoral (i.e. religious). They
47
are both religious and moral. Thus, there can be a religious qualification of
morality or a theological qualification of ethics. It is also possible to interpret
the criterion of universalizability so that religious action-guides also qualify
as moral ones.
One objection to taking formal features of morality as sufficient rather
than merely necessary is that they undercut or make it difficult to explicate
our ordinary distinctions between morality and law, religion, etc. But, given
the view that moral judgments must meet a social material condition, can religious
action-guides meet this condition? When H bases his argument on- God ' s will,
isn't he thereby excluding all reference to human welfare or harm? Denying that
the condition of concern for human welfare or harm writes "humanism" or "secularism"
into the definition of morality, G. J, Warnock adds,
I suspect that religious views differ from 'humanist' views,
not by denying the essential moral relevance of human benefit
or harm, but rather by incorporating very different beliefs
as to what really is good or bad for human beings. The re-
ligious believer finds in a supernatural order a whole extra
dimension of pre-eminently important gains and losses, benefits
and harms; his difference with the non-believer is not on the
question whether these are of moral significance, but simply
on the question whether they are real or chimerical. He might
also wish to expand what; might be called the moral population
to include moral beings supposed not to be human..., AS
Or "moralities with a transcendental dimension have a greater logical freedom
with respect to their content than moralities which h;.ve no such dimension..,."'
7-27
To take an extreme statement, one might hold on some religious views that
it is right to destroy another's body in order to save his soul. If such a
statement meets the formal conditions of morality, it may also meet the social
material condition. Religious viewpoints then may q uail f y morality by offering
a different set of facts or a different interpretation of the same facts regarding
human welfare. This contention would also hold for other worldviews or large
pictures of life, for instance, Marxism, which single out one benefit or harm
(such as economic deprivation) or could give preference to some beings (such as future
generations) over others. In such interpretations, metaphysical or theological
convictions provide insights into humanity and its place in the universe without,
however, generating any new basic moral principles, which may be sufficiently
clear by ordinary experience or reason. Nevertheless, these "evaluative de-
scriptions" of facts enable us to infer "new derivative duties from basic ones
already known to us." For instance, we may acknowledge a duty of beneficence or,
at least, nonmalef icence, but our conception of specific duties will depend on
our interpretation of human welfare and harm.
Some religious views on abortion do not hinge on the revelation of a novel
basic moral principle but rather on what constitutes human welfare, the weight
given to some harms in relation to others, and, especially, when human life begins.
A basic moral principle is that we are not permitted to kill other human beings
except under well-defined conditions; this principle is shared by most moralities,
religious or secular. If one adds an "evaluative description" of the fetus as
human being from the moment of conception, the prohibition of abortion follows.
This expansion of the population about whom one must be concerned can be construed
as religious insofar as one of its premises is held on the grounds of religious
authority or on grounds about the ''beginning of human life that are not merely
empirical (but then no statement that human 1 if c begins at this or that point in
time can be merely empirical). Some recont moral attacks on the prohibition of
abortions have accepted this religious- Vi-rsi on of the beginning of human life for
7-28
sake of argument and have then tried to show that the relevant basic moral
principle is not non-maleficence but rather beneficence. The question then
becomes "how much ought we to do to sustain human life?" rather than "when may
we kill human beings?" But even here a religious position might affirm an
interpretation of beneficence along the lines of self-sacrifice, going the second
mile, etc.
Another interpretation of the relation of religion and morality would take
a more stringent view of the criterion of otherregardingncss insisting that since
human welfare and harm have relatively determinate meanings a religious action-guide
must address those meanings if it is also to be moral. From this standpoint, the
concept of human welfare has a relatively fixed core "which includes such 'objective'
conditions as physical survival, bodily and psychic health, security from arbitrary
violence, and the like." Little and Twiss contend that when a religious action-
guide calls for a violation of this core of human welfare, perhaps on the grounds
that the other's welfare is "really" or "truly" different from these objective con-
ditions, it cannot be considered "moral,"
But consider variations of the Abraham and Isaac story. Abraham's conviction
that he should sacrifice his son, even if he conceived this to be truly in his son's
best Interests, would not be "moral" but only "religious" given the Little-Twiss
condition. But what if Abraham decided to sacrifice his son because God had
promised to make the people "happy" if he performed this act? Little and Twiss
presumably would deny that this is a moral argument since survival is part of the
core concept of human welfare whereas happiness is not. But suppose that Abraham
felt that God had commanded this sacrifice as expiation for his people's sins and
that he should perform it in order to stop Cod from slaughtering several thousand
people. Here Abraham does not disregard Isaac's physical welfare for the sake of
Isaac's "real" wel fare;" nor does he sacrifice Isaac's physical welfare for the
sake of others' happiness. Ralher, Abraham balances Isaac's death against the
7-29
deaths of several thousand persons. Is this to be a "moral" decision or a
"religious" one? Does the presence of a "transempirical" premise (that God will
do what he said he will do) render this a "religious" argument? Would it be a
mixed "religious" and "moral" position?
One possibility is that we set certain standards of evidence regarding
causal processes, etc., and do not consider positions that fall short of those
standards as moral (at least in a critical sense although they may be mora] in
an anthropological sense). For, as Ronald Dworkin contends,
If I base my position on a proposition of fact ( ' homosexual
acts are physically debilitating') which is not only false,
but is so implausible that it challenges the minimal standards
of evidence and argument I generally accept and impose upon
others, then you would regard my belief, even though sincere,
as a form of rationalization, and disqualify my reason on that
ground. 52
Perhaps some religious beliefs, especially if they contradict basic assumptions
about human welfare, may be disqualified in the same way. But if we work from
this perspective, we are clearly offering more than a neutral classification of
reasons and statements. Indeed, the attempt to draw some sharp distinctions
between "religious" and "moral" reasons may presuppose a value-judgment to the
effect that religious reasons for policies must be viewed with suspicion and
perhaps even disqualified at least in a society marked by a sharp separation of
church and state and religious pluralism. Here, again, what is actually an
issue of how much power and influence certain sorts of beliefs should have in a
society is transmuted into a putatively neutral, logical analysis of terras.
7-30
VII Conclusion
Because of space requirements, I am unable to Include a projected section
on legal action-guides and considerations, which would have considered issues
raised by F and G and examined the logical and conceptual relations between law
and morality. It would have treated two major approaches in jurisprudence as
analogous to the two major options in defining morality: formal conditions and
material conditions. Those who insist on a moral content in the definition of
law (e.g. "an unjust law Is no law at all") stand In opposition to those who de-
fine law In more formal terms of commands, rules, etc. although there are several
mediating interpretations between "natural law" and "positivism" (two widely used
but unhelpful labels) .
I have tried to highlight several issues in the identification of "ethical"
or "moral" judgments, reasons, and action-guides, concentrating on three formal
conditions and one social material condition, and I have dealt with the distinction
between "moral" and "nonmoral" in relation to ideals, prudence, and political and
religious considerations. Although I have been mainly concerned with analysis, I
think that a good case can be made that the conditions of prescrlpti vi ty , univer-
salizabi lity , and otherrcgardingness are necessary and sufficient conditions of
moral Judgments and action-guides, but I concentrated on the implications of taking
either a formal or a material definition for such topics as ideals and prudence. I
did, however, explicitly exclude finality or overridingness , at least in its strong
or absolute sense, from the criteria of "moral", mainly because it begs too many
questions, particularly about whether one should be moral and what wei ght moral
reasons should have in public debate. Attempts to classify reasons as "moral" and
"nonmoral" are not too misleading and dangerous if the classifiers do not assume
that "moral" reasons are ultimately authoritative or are the final arbiters. Moral
reasons may well have this place but definitions should nut foreclose such important
human and public questions. Thus, even if basically prudential or egoistic reasons
7-31
are nonmora] , their function is akin to moral ac t ion-guides , and the distinction'
between moral and nonmoral should not prejudge their weight.
In some social and political conflicts one party may have the power to
punish or restrain the other. Consider the case of government X and resister A,
whose position that the war is wrong satisfies only the formal criteria of
morality (i.e. avoids consideration of effects on others) ■ I'rankenn insists that
the formal definition of morality "makes it unclear just w hat moral just If i cat ion
society could have for punishing or restraining any individual who is acting
'conscientiously' in accordance witli his own AC [action-guide], whatever this may
53
be and whatever his action is." We have already seen that this dispute could be
called moral without attributing moral claims to both sides, but, given only formal
criteria, both sides are playing the moral game. The moral justification for the
government's restraint and for A's resistance (flight, etc,) would come from
substantive moral positions. We are only concerned with whether the reasons can
be construed as moral not whether they can be actually justified on substantive
grounds. Furthermore, it is not at all clear that insisting on a social material
condition of morality in any way avoids this issue. For a dispute between government
X and resister B or C, both of whom consider effects on others, is still a dispute
within morality, between two substantive moral positions, X may restrain or punish
B or C if they disobey or flee on the grounds that their acts are "morally un-
justified" although their reasons are recognized as moral ones (as opposed to non-
moral). A definition of "moral" over against "nonmoral" cannot resolve such disputes
or even make them easier to resolve. Frankena's objection appears to trade on the
ambiguity of "moral," at this point taking it as opposed to "immoral" instead of
"nonmoral." Indeed, his objection hallows the liberal notion that there can be
no moral j ustificatio n for the state's interference in conduct unless there is harm
or the imminent danger of harm to others or to society. This position may well be
true (and I am Inclined to think that it Js, with Homo qualifications), but it
7-32
should not be declared true by the definition of "moral." Tt is a matter for
substanti ve debate.
One major concern about all the definitions we have examined is that certain
sorts of reasons- may receive n great dea] of power and influence by virtue of
their classification as "moral" over against "nonmoral." An allegedly neutral,
logical analysis may serve as a .substitute for substantive debate, not only setting
the terms of the debate but also determining the weight of sorts of reasons.
"'When I use a word,' llumpty Dumpty said, in rather
a scornful tone, 'it means' just what I choose it to mean-
neither more nor less,
'The question is,' said Alice, 'whether you can make
words mean so many different things,'
'The question is,' said llumpty Dumpty, ' which is to
be master — that's all.'"
7-33
Notes
1. In tliis paper I use a variety of expressions such as considerations,
reasons, and action-guides as shorthand for a range of nouns such as arguments,
rules, principles, etc. that might be modified by "ethical" or "moral."
2. See Parsons, "Christian! ty and Modern Industrial Society," in Socio! o gi cal
Theory and Modern Society (New York: The Free Press, 3967); most of the essay is
reprinted in James F. Childress and David 11. llarned, eds. Secularisation and tjie
Protestant Prospect (Philadelphia: The Westminster Press, 1970), pp. 41-70.
3. Thomas Luckmann, The Invisib le Religion (New York: The Macmillan Co., 1967),
p. 101.
4. I shall use the terms "moral" and "ethical" interchangeably in this dis-
cussion although one might wish to distinguish them in some contexts. Cicero
apparently formed the Latin word moralis (from mores) to translate the Greek term
ethikos . Etymologically their meanings are very similar, stressing manners, character,
and customs. Contemporary usage suggests some rough hut not very precise distinctions
between them. "Ethics" often expresses more intellectual perspectives or even
theories, while "morality" often refers more directly to practice and actual conduct.
Recent philosophical discussion has been focused on the concept of "morality" (not
"ethics"). Likewise, moral (rarely "ethical") principles and arguments have been
the subject of discussion.
5. William K. Frankena, "The Concept of Morality," in The De f inition of
M orality , edited by A.D.M. Walker and C. Wallace (London: Methuen, 1970), p. 149.
Frankena uses the terms "descriptive-elucidatory" and "normative," the latter of
which I avoid because of its use also in normative ethics, although 1 recognize
that not all normative judgments are moral ones. Frankena's approach is clearly
one of normative metaethics, and he at times appears to offer normative moral reasons
for his distinction between moral and nonmoral.
6. See R. M. Hare, Fr eedom an d Reason (london: Oxford University Press, 1963),
pp. 146 ff. Contrast William K. Frankena, "The Concept of Morality," The Def ini t ion
of Morality , pp. 160-161. See my discussion in Part III below. Much of my dis-
cussion in this paper has been influenced by a number of Frankena's essays.
7. Joel Feinberg, Doing and Deserving: Essays in the Theory of Responsibility
(Princeton, N.J.: Princeton University Press, 1970), p. 18, fn. 10.
8. W.D. Hudson, Modern Moral P hilosophy (Garden City, N.Y.: Doubleday, Anchor
Rooks, 1970), p. 2.
9. These four criteria which appear in various forms in the literature have
been thoroughly discussed by William Frankena in several places (for instance,
"The Concept of Morality" in The Def ini t ion of Morality) . Other philosophers have
identified other criteria although I am inclined to think that they are subordinate
to these four. Thus G. Wallace and A.D.M. Walker consider definitions in terms of
particular forms of sanct ions that are associated with moral rules and principles,
and in terms of the importance of moral rules and principles. It is not at all clear
that importance can be distinguished from overridingnnss , unless one is thinking about
importance for society. Not Is it clear that a discussion of sanctions, either in-
ternal or external, would he useful, [or sanct i ons del ive from morality's author i -
tati veness . See Wallace and Walker, eds., Tile l)cf_in_i. I ion of Mora l I ty, "Introduction,"
7-34
Also in that volume, see T.L.S. Sprigge, "Definition of a Moral Judgment," which"
emphasizes sanctions, and Neil Cooper, "Morality and Importance."
10. R. M. Hare in Freedom anil Reason appears to defend all three formal criteria,
especially the first two, while explicitly rejecting the material one. By taking
"prescriptivity" in a limited and weak sense, I am not doing Justice to Hare's
position, although this is sufficient for our purposes,
. 11. Hare, Fre edom and Reas_on, pp. 139-40.
12. Ibid. , pp. 168-69.
13. Ibid ., p. 195.
14. See William K. Frankena, "The Concept of Morality," The Jou rnal of
Philosophy , I.XXIII (1966), reprinted in Kenneth Pahel and Marvin Schiller, eds.,
Readings in Contemporary Ethical Theory (Englewood Cliffs, N.J.: Prentice-Hall,
1970), p. 394. Cf. also the longer earlier version of this paper with the same
title, "The Concept of Morality," The Definition of Morality .
15. See Alasdair Maclntyre, "What Morality Is Not," Philosophy (1957) : 325-35,
Reprinted in The Definition of Morality , pp. 26-39.
16. See William K. Frankena, "The Concept of Morality," The Definitio n of
Morality , p. 154.
17. G. J. Warnock, Contemporary Moral Phi losoph y (New York: St. Martin's
Press, 1967), p. 54. Several of Frankena's essays discuss and propose the social
material criterion. Other philosophers who have proposed versions of this criterion
include Kurt Baier, Bernard Williams, J. Kemp, David Cauthier, Stephen Toulmin, and
Philippa Foot. Among the proponents of some version of formal conditions of morality
are R. M. Hare, H.L.A. Hart, W. D, Falk, and John Ladd.
18. Warnock, Contemporary Morn] Phi J o s oph y , p. 69, cf. pp. 60-61. Although
these are matters of judgment, they are not matters of moral judgment, and Warnock
insists that the charge of circularity does not apply when one then defines morality
in terms of these "facts." In this paper I cannot, of course, deal with the debate
about whether this approach can avoid or rebut the charges leveled against "naturalism,'
which is at the heart of the debate,
19. See David Little and Sumner B. Twiss, "Basic Terms in the Study of Religious
Ethics," Religion and Morality , edited by Gene Outka and John P. Reeder, Jr. (Garden
City, N.Y.: Anchor Press/Doubleday , Anchor Books, 1973), pp." 54 ff.
20. "The Concept of Morality," The Definition of Morality, p. 156, Emphasis
added.
21. Ibid. , p. 157,
22. See Hare's charge in Freedom and Reason, p, 163. Contrast Frankena,
"The Concept of Morality," Readings in Contem porary Krhical The ory , pp. 396, 394,
It is interesting that in norm a t I vo ethics, a number of different parties to the
dispute are utilitarians or mixed theorists. But these normative ethical positions
bear no close relation to their views of the nature of morality. Regarding dis-
tributive justice, Hare writes, "I do n< i think that the difference between
naturalism [which, in this context, would include the social material definitions
7-35
of morality] and prescriptivism [bis own position] has any bearing on the
matter — which is hardly surprising if, as I think, there is no substantial
difference between the parties on ajry_ matter which is going to affect our actual
moral arguments." "Wrongness and Harm," Essays on the Mora l Co ncepts (Berkeley
and Los Angeles: University of California Press, 1972), p. 96,
23. Another possibility is to distinguish between "strong" and "weak" senses
of over'ridingness or priority. Strong priority means mora] action-guides always
take precedence over conflicting nonmoral action-guides; weak priority indicates
a prima facie priority so that mora] action-guides cannot always be ignored or over-
ridden without ceasing to be moral. Little and Twiss develop this interpretation
and defense of weak priority in "Basic Terms in the Study of Religious Ethics,"
Religion and Morality , p. 48. I would also emphasize that weak pr ima facie priority
means that moral action-guides must always be taken seriously; they cannot be
disregarded although they may be overridden. Furthermore, one might extend the
analysis to indicate presumptions, burdens of proof, etc. Although one might dis-
tinguish finality or ultimacy and priority, I am conflating them in my discussion,
24. See Strawson, "Social Morality and Individual Ideal," in Readings in
Contemporary Ethica l Theory , edited by Kenneth Pahel and Marvin Schiller, pp. 344-59,
Although space limits prevent me from pursuing the point in this paper, frequently
one does not try to persuade the others that they ought to adopt one's ideals, which
are purely persona] or rest on religious and other beliefs. In the debate between
government X and resister E, the latter would probably ask for a personal exemption
rather than trying to persuade X that conscription itself is wrong.
25. See Hare, Freedom and Reason , pp , 146ff.; W, D. Falk, "Morality, Self, and
Others," in Morality and the La nguage of Conduct , edited by Hector-Neri Casteneda
and George Nakhnikian (Detroit: Wayne State University Press, 1963), pp. 25-67;
Frankena, "The Concept of Morality," The Definition of Morality , pp, 159-163,
26. This option, which I find unsatisfactory, is similar to Frankena's strategy
in his discussion of Hare's example of a debate about ideals in relation to a "stripper'
in a club catering to middle-aged businessmen, "The Concept of Morality," The
Definition of Morality, p. 160; cf . Hare, Freedom and Reason , pp. 147f.
27. Warnock, Contemporary Moral Philosophy , p. 59.
28. Cf. W. D. Falk in "Morality, Self, and Others," in Morality and the
L anguage of Conduct .
29. Contrast Hare who thinks that such an action-guide may be prescriptive
(and perhaps taken as overriding) without being universalizable. He distinguishes
"(which meets both formal and social material conditions), "ought2" (which
is prescriptive and universalizable), and "oughtn" (which is prescriptive but not
universalizable). Although Hare does not think we actua]]y ever use "ought" in the
third sense, he considers it: "'Oughts' lias to do with questions of self-interest
whicli is not universalized-self-interest , and the interest of groups, such as my_
family, and my_ country, which are defined by reference to an individual," Freedom
and Reason , p. 165.
30. See "The Concept of Morality," The Definit ion of Morality , p, 156,
7-36
31'. Wnrnock, Contem porary Moral Philosophy, p, 49, which summarizes Lhe
position taken by Baler and CauLhier, among others, as well as his own.
32. Josef Pieper, Prudence. Translated from the German by Richard and Clara
Winston (New York: Pantheon Books, 1959), pp. 21-22, 53.
33. Cf. W, D. KAlk's perceptive comment: "This was, In effect, the view
of the old Natural Law moralists — Hooker, Grotius, I'uf f endorf : Lhe social virtues
derive joint support from our natural concern for our own good and for that of
society. Hohhes streamlined this account by denying Lhe second, which provoked
subsequent moral isLs to deny the first. UoLh llobbes's sophistical toughness and Lhe
well-bred innocence of the academic moralists since are distorted visions which
are less convincing than the unsqueaml sli common sense of the philosophers and
divines of earlier times." "Morality, Self, and Others," in Morality and the
Language of Conduct.
34. This paragraph derives from 1'alk's account, see ibid .
35. Giovanni Sartori, "What is 'Politics'?" Political Theory 1, no. 1
(February 1973): 17-18, Contrast J. R. Lucas who uses the term "political" in
its undifferentiated, wide, original sense "in which it applies not only to politics
in the modern, narrow, sense, but to social and. legal affairs, and all that pertains
to men's public life." The Principles of Politics (London: Oxford University Press,
1966), p. 20, fn. 1.
36. Virginia Held, "Justification: Legal and Political," Ethics 86 (October 1975)
37. Hanna Pitkin, Wittgenstein and Ju stice (Berkeley and Los Angeles: University
of California Press, 1972), p. 215; Bernard Crick, In Defense of Politics (Baltimore:
Penguin Books, Revised Pelican Edition, 1968), p. 148.
38. W. D. Falk, "Morality, Self, and Others," Morality and t he Language of
Conduct , p. 39.
39. Gewirth, "Political Justice," Soci al Justice, edited by Richard Brandt
(Englewood Cliffs, N.J.: Prentice-Hall, 1962), pp. 119-20.
40. This discussion of ordinary usage stems largely from Hanna Pitkin,
Wittgenstein and Justice , pp. 214-15.
41. Weber, "Politics as a Vocation," in From Max Weber : Essays in Sociology ,
ed. and trans. Hans Gerth and C. Wright Mills (New York: Oxford University Press,
1946), pp. 118ff.
42. Virginia Held seems to claim too much about the teleological characler
of political reasoning and the deontological character of legal reasoning although
some of her points certainly hold. "Justification: legal and Political," Ethics 86
(October 1975) : 1-16.
43. Both religion and morality, which are usually paired, relate more clearly
to practice, while both theology and ethics imply more reflection upon prat Lice,
James Gustafson writes, "Theology is reflection on human experience with reference
to a particular dimension of the human experience donated 'religious,'" The
Contributions of Theology to Medical I'.Liiies. The 197 r > Pere MarqucLlc Theology
Lecture (Theology IVpl., Marquette University, Apr I I l l! 7S), p. 4, cf, p. 6-7. In
these remarks 1 shall concentrate on del' i n 1 1 Ions of "religion," assuming that
7-37
theology is reflection on "religious" objects, experiences, etc,
i
14, Tillich, Christianity and the Encoun ter of the WorU Religions (New York:
)ia University Press, 1963), p. 4; cf. Dyn amics of Fait h (New York: Harper
44,
Columbi
and Row, 1958), pp. 1-12. See the good critique of Tillich's definition in
Rem B, Edwards, Reason and Religion (new York: llarcourt. Brace, Jovanovich, 1972),
pp. X 3-13.
45.' Little and Twiss, "Basic Terms in the .Study of Religious Ethics,"
R el i g I on and Mo r a 1 i t y , p . 62.
46. Frankcna, "Conversations with Carney and Hauerwas," The Journal of
Religious Ethics , III (Spring 1975) : 56, Both Fred Carney and James Gustafson call
into question the universal: zability requirement for "moral" in order to make room
for "religious ethics." See Carney, "On Frankena and Religious Ethics," The Jou r nal
of Religious Ethics , III (Spring 1975): 7-25 and James M. Gustafson, Can Ethics Be
Christian ? (Chicago: University of Chicago Press, 1975), passim, I think that
Frankena 's point holds against both positions as far as judgments are concerned,
47. Gustafson, Can Ethics Be Christian?, p, 177: "On the basis of the assumption
that certain values and principles have an obligatory character within a 'way of life'
and that the Christian history and community call for a way of life grounded in the
Christian story, it is fitting to call these Chri stian ethical principles ant[ values , "
My emphasis. Cf. p. 167: "It is a moral principle in the sense that it determines
the 'conscience' of the believer...," But Gustafson also writes (p. 117), "It
[a principle of nonviolence] is a moral principle because actions governed by it have
moral, or at least human value consequences for others," It is not clear whether he
considers this social material concern or effect to be a necessary or sufficient
condition.
48. Warnock, Contemp orary Moral Philosophy , p. 79, fn. 27.
49. Bernard Williams, Morality : An Introductio n to E thics (New York: Harper
and Row, Harper Torchbooks, 1972), p„ 84.
50. William K. Frankena, "Is Morality Logically Dependent on Religion?"
Rel igio n and Morality, edited by Gene Outka and John P. Reeder. I cannot deal
with a number of the important issues surrounding the "is -ought" debate,
51. David Little and Sumner B. Twiss, Jr., "Basic Terms in the Study of
Religious Ethics," Religion and Morality , p. 74.
52. Ronald Dworkin, "Lord Devlin and the Enforcement of Morals," Moral i ty
and the Law , edited by Richard A. Wasserstrom (BelmonL, Cal.: Wadsworth Publishing
Co., 1971), pp. 63-64.
53. "The Concept of Morality," The Def inition of Morality , p. 159. My
emphasis.
7-38
8
BASIC ETHICAL PRINCIPLES IN THE CONDUCT OF BIOMEDICAL AND
BEHAVIORAL RESEARCH INVOLVING HUMAN SUBJECTS
H. Tristram Engelhardt, Jr., Ph.D., M.D.
I. INTRODUCTION
Philosophy serves us best by helping us be clearer about the meaning
of activities such as knowing and valuing. Philosophy does not simply map
actual human activities in conceptual terms. Rather, it provides a por-
trayal of the conceptual topography of the possible. Thus, when individuals
inquire after the basic ethical principles which should underlie the conduct
of biomedical and behavioral research, they are not (or should not) be ask-
ing what principles actually do underlie such research, or even what prin-
ciples we would be most comfortable with. After all, our conduct may be
influenced by a complex of prejudices which would incline us to do things
which we would, after more penetrating reflection refrain from doing. This
is not to say that ethics deduces guidelines for human conduct from some
divine perspective. Surely not. Ethics, as a philosophical enterprise, is
best conceived as an attempt to negotiate diverse moral intuitions. Ethics
is the logic of a pluralism in the sense that ethics is an attempt to find
the most general grounds or bases for judging the rightness and wrongness
of conduct. Unlike religious ethics, or particular legal traditions, philo-
sophical ethics hopes for general principles of conduct discoverable by dis-
interested reflection, apart from either grace or cultural prejudice. Though
such a disinterested perspective cannot be attained, one can move towards
such a vantage point by attempting to lay out ever more clearly general prin-
ciples of moral conduct. It can thus function as a regulative ideal in the
attainment of greater objectivity in moral judgments, in the sense of striv-
ing for greater intersubjectivity ( i.e. , principles more likely to enjoy
general assent).
What I will provide here is an analysis of what is at stake with regard
to moral conduct in the case of biomedical and behavioral research involving
human subjects. To begin, I will hold that in general there are two central
and different types of concerns at issue in experimentation involving human
subjects: (1) questions bearing on respect for persons as moral agents, and
(2) questions bearing on the achievement of goods and values which humans
in general esteem. The first set of considerations turns on respect for
the freedom of individuals. The second set of considerations turns on the
best interests of such individuals, and of the human community in general.
Most people will readily agree that there are goods and values which
humans esteem. There are, though, some individuals who hold that freedom
2
is a non-scientific chimerical concept. Since appeals to freedom in con-
cepts of free and informed consent are so important to the ethical issues
in experimentation, it is worth indicating here the central role of the
concept of freedom for morality. Holding that persons are free is not a
metaphysical thesis, but the statement of a proposition that must be pre-
3
supposed if any talk about human conduct is to make sense. If persons
are not free to choose good reasons over bad, but are caused to embrace
some reasons rather than others, then there is no possibility of anyone
being right or wrong about anything--and no need for this essay. For in
that case, persons would simply be caused to hold that some activities
were more or less successful, correct or incorrect, valid or invalid. To
be right or wrong about matters of fact or logic presupposes that one could
hold certain propositions to be true or valid on the basis of good reasons,
rather than being caused to assert that they are true or valid.
8-2
The same holds for moral conduct. One must presuppose that persons can
freely choose to do one thing rather than another if one is to mean anything
else by morally correct (or incorrect) conduct than that one is reinforced
(or negatively reinforced) by such conduct, or the prospect of such conduct,
and that one is reinforced by others engaging in (or not engaging in) such
conduct. Morality presupposes that individuals are worthy of blame or praise
because they can freely choose between different lines of conduct. If that
is not the case, then we are simply caused to engage in particular behavior
and to call certain behavior moral or immoral, and there is no possibility
to mean anything by 'right' or 'wrong' other than that one is caused to call
some things right or wrong. In that case, any serious talk of ethical prin-
ciples must cease. For example, the proposition, "Nazi experiments on non-
consenting human subjects were evil," would mean, in the absence of the pre-
supposition of human freedom, only that one was negatively reinforced by
the thought of what the investigators did, and was positively reinforced by
condemning that conduct. It would not follow that such investigators were
actually right or wrong, but simply that one had been caused to feel dis-
tressted on account of such conduct. One could choose between those two
accounts of human action only if one were able to choose a good account
over a bad account—which is to say, if one is free. It is in this sense
that the presumption of freedom is a necessary presumption.
This analysis of basic ethical principles concerning experimentation
with human subjects begins then with the presupposition that there is a
sense of moral responsibility, and that this sense has meaning only if per-
sons think of each other as being free. Thinking of one another as free is
8-3
a logical condition for the existence of a moral community, a community bound
together on the basis of mutual respect, rather than force. This is the key
consideration in the requirement of free and informed consent—consider, for
example, the first requirement of the Nuremberg Code, "The voluntary consent
5
of the human subject is absolutely essential." Of course, treating persons
as members of a moral community (in the case of experimentation, requiring
voluntary consent) requires that the members of that community ( e.g. , the
human subjects involved in experimentation) can indeed be treated as free
agents— a condition not met, for example, by very young children or many
of the very senile, mentally ill, or mentally retarded. In short, not all
human subjects can be treated in the same way with regard to ethical consi-
derations in experimentation involving human subjects, because such subjects
differ in at least one very relevant consideration, the ability to choose as
free agents.
To lay out the basic ethical principles involved in human experimenta-
tion will then require: (1) indicating what is due to human subjects used
in research out of respect for them as free moral agents; (2) indicating
what is due to those human subjects who cannot be treated as free moral
agents; and (3) outlining the basic values and goods we wish to support
in the practice of human experimentation. In all of this, if one is to
speak of basic ethical principles, one must distinguish those principles
from procedures or guidelines employed in safeguarding such principles.
Thus, requiring free consent is a basic ethical principle, while requiring
written consent would merely be a procedural safeguard.
8-4
II. A SKETCH OF BASIC PRINCIPLES
The literature concerning human experimentation is rich and varied.
It is for the most part concerned with establishing procedures and safeguards
in the use of human subjects, rather than providing an analysis of, or clear
presentation of, the ethical principles at stake. This lack of clarity stems
in part from the fact that any one of the basic procedures involved in en-
suring the moral use of human subjects in experimentation involves more than
one basic ethical principle. I will hold in the course of this analysis that
the literature focuses on three cardinal ethical issues. The first concerns
respect for persons as a logical condition for morality. Such respect for
persons is not a value among other values. It is rather the basis for our
sense of moral responsibility, and is considered apart from any interest we
might have in respecting other persons ( e.g. , that such respect is useful,
or that giving such respect will tend to protect us). It is a concern for
o
rights. The second and third are concerns for goods and are thus teleolo-
Q
gical— or goal -oriented. These ethical principles are:
A. One should respect human subjects as free agents out of
a duty to such subjects to acknowledge their right to
respect as free agents.
B. One should foster the best interests of individual human
subjects.
C. One should have concern to maximize the benefits accruable
to society from research involving human subjects, taking
into particular regard interest in values such as (1) the
8-5
amelioration of the human condition through advances in
the biomedical and behavioral sciences and technologies;
(2) preservation of human autonomy as a general value;
(3) increase in knowledge apart from any consideration
of its application to the amelioration of human condi-
tion; (4) the personal satisfaction of human subjects
derived from their feeling of having contributed to the
common good or to the advancement of human knowledge
by participation in research.
The first principle is a deontological one in the sense of focusing on
a consideration of rights and duties independently of any issue of goods and
values. It is an appeal to respect for the freedom of persons whether or not
such respect would in the long run contribute to the benefit of society.
With regard to this principle, experimentation upon unwilling human subjects
should be regarded as immoral, even if the results of such experimentation
would be of considerable general utility. It would be with regard to such
experimentation that the Nazi use of human subjects would be worthy of con-
demnation, even if it had been the case that such experimentation had revealed
extremely useful information not otherwise attainable. Such basic rights can-
1 2
not be outweighed by goods.
The second principle, concern to foster the best interests of the indi-
viduals involved, is a particular type of teleological moral concern, which
may or may not be utilitarian in its nature. One may be concerned with the
best interests of others (1) out of feelings of sympathy or fellowship, (2)
8-6
out of interest in maintaining a society which would act fairly by allowing
only that experimentation which redounded to the advantage of the least well-
off member of society (and therefore to the best interests of all members
on the supposition that a cardinal interest of persons is to avoid the fear
of possible exploitation), (3) out of fear that if one did not as a general
rule support the best interests of others, one's own best interests would
be set in jeopardy, etc. One should notice, though, concern for the best
interests of others may not coincide with respecting them as free agents.
Free agents presumably can act against their best interests. Therefore,
concern for the best interests of others may lead to certain acts of pater-
nalism in which one restricts the freedom of individuals in order to support
their best interests. One might think here of rules circumscribing the oppor-
tunity of prisoners to volunteer as subjects of human research in that such
a liberty is not in their best interests (cf. laws which forbid persons the
1 o
freedom of selling themselves into slavery).
The third moral principle is as well an axiological or teleological
principle. It does not turn on a recognition of a basic duty to perform such
actions, but rather on an interest in certain goods or values. This interest
can be construed within some formualtion of the utilitarian principle of maxi-
mizing the greatest good for the greatest number. It is within the context
of such considerations that calculations most naturally arise concerning the
cost/benefit ratio of particular experiments involving human subjects. The
considerations ( i.e. , goods and values) usually at stake in such calculations
include (1) interest in ameliorating the human condition through advance in
biomedical and behavioral sciences and technologies, (2) preserving a general
social interest in individual autonomy, (3) supporting the advance of know-
ledge, and (4) allowing individual humans the satisfaction of contributing
to the general good.
It is worth noting that interest in autonomy here is different from the
concern to respect persons as moral agents. Here human freedom is considered
as one value among others. Thus, it is one thing for an experimentor to gain
free consent from a volunteer for participation in an experiment, and another
to encourage that subjects consider carefully their consent so as to maximize
their opportunity to act freely. In the first principle, resepct of free-
dom functions as a limiting criterion for moral conduct; in this case respect
for freedom functions as a value around which to structure a society. It
has a utility value. This focus on freedom as a general value for society
is a teleological or axiological moral concern. Framed in this context,
the "right to autonomy" expresses a general social interest in the value of
autonomy. The right to autonomy is thus reducible to talk about interest
in values or goods; talking of a "right" here is a sub rosa way of enjoin-
ing the pursuit of a value.
The three moral principles of (1) respect for persons as free moral
agents, (2) concern to support the best interests of human subjects in re-
search, and (3) interest in assuring that the use of human subjects in ex-
perimentation will on the sum redound to the benefit of society, appear
in various fashions in the literature and codes concerning human experi-
mentation. When comparing the Nuremberg Code, the Declaration of Helsinki,
the Department of Health, Education and Welfare rules and regulations for
the Protection of Human Subjects, and the Ethical Principles in the Conduct
of Research with Human Participants of the American Psychological Associa-
tion, one finds each of these three principles, though in various and often
mixed forms.
A. Respect for human subjects as free agents:
The Nuremberg Code: "1. The voluntary consent of the
14
human subject is absolutely essential.'
Declaration of Helsinki: "III. 3a. Clinical research on
a human being cannot be undertaken without his free con-
1 5
sent after he has been informed."
DHEW Regulations: '"Informed consent' means the knowing
consent of an individual or his legally authorized repre-
sentative, so situated as to be able to exercise free
power of choice without undue inducement or any element
of force, fraud, deceit, duress, or other form of con-
straint or coercion."
American Psychological Association: "3. Ethical practice
requires the investigator to inform the participant of all
features of the research that reasonably might be expected
to influence willingness to participate and to explain all
other aspects of the research about which the participant
inquires. ... 4. Openness and honesty are essential charac-
teristics of the relationship between investigator and re-
search participant. ... 5. Ethical research practice re-
quires the investigator to respect the individual's freedom
8-9
to decline to participate in research or to discontinue
participation at any time."
B. Concern to foster the best interest of the individual
human subjects:
The Nuremberg Code: "4. The experiment should be so
conducted as to avoid all unnecessary physical and mental
suffering and injury. 5. No experiment should be con-
ducted where there is a priori reason to believe that
death or disabling injury will occur; except, perhaps,
in those experiments where the experimental physicians
also serve as subject. ... 7. Proper preparations should
be made and adequate facilities provided to protect the
experimental subject against even remote possibilities
of injury, disability, or death. ... 10. During the
course of the experiment the scientist in charge must
be prepared to terminate the experiment at any stage,
if he has probable cause to believe, in the exercise of
the good faith, superior skill, and careful judgment
required of him that a continuation of the experiment
is likely to result in injury, disability, or death to
I Q
the experimental patient."
Declaration of Helsinki: "II. 2. The doctor can combine
clinical research with professional care, the objective
being the acquisition of new medical knowledge, only to
the extent that clinical research is justified by its
8-10
19
therapeutic value for the patient." "In the purely
scientific application of clinical research carried out
on a human being it is the duty of the doctor to remain
the protector of the life and health of that person on
20
whom clinical research is being carried out."
DHEW Regulations: "It is essential that the committee,
representing a wide spectrum of those expert professional
skills essential to a clear recognition of an activity's
inherent risks and probable benefits, carefully weigh
such risks and benefits before determining that the bene-
fits favor a decision to allow the subject to accept these
risks." (My emphasis). This implies that volunteers
should not be allowed to choose what the committee held
was not in their best interests.
American Psychological Association: "7. The ethical in-
vestigator protects participants from physical and mental
discomfort, harm, and danger. If the risk of such conse-
quences exists, the investigator is required to inform the
participant of that fact, secure consent before proceeding,
and take all possible measures to minimize distress. A
research procedure may not be used if it is likely to cause
22
serious and lasting harm to participants."
C. Concern to maximize the benefits accruable to society from
research involving human subjects:
The Nuremberg Code: "2. The experiment should be such
8-11
as to yield fruitful results for the good of society, un-
procurable by other methods or means of study, and not
random and unnecessary in nature. ... 6. The degree of
risk to be taken should never exceed that determined by
the humanitarian importance of the problem to be solved
23
by the experiment."
Declaration of Helsinki: "I. 3. Clinical research cannot
legitimately be carried out unless the importance of the
objective is in proportion to the inherent risk to the sub-
ject. 4. Every clinical research project should be pre-
ceded by careful assessment of inherent risks in comparison
24
to foreseeable benefits to the subject or to others."
DHEW Regulations: "The risks to the subject [must be]
outweighed by the sum of the benefit to the subject and
the importance of the knowledge to be gained as to war-
rant a decision to allow the subject to accept these
risks." (My emphasis).
American Psychological Association: "Where scientific
or humane values justify delaying or withholding informa-
tion the investigator acquires a special responsibility
to assure that there are no damaging consequences for
the participant."
Even in the above excerpts which are chosen to highlight each princi-
ple in turn, there is often a simultaneous appeal to one or both of the
other two principles. This must stem in part from the ways it is in fact
8-12
useful to establish procedural maxims for ethical conduct in human experi-
mentation. For example, free and informed consent functions not only as
a means of ensuring respect for persons, but as a way to pursue other values
such as having a society in which one would not live in fear. It is there-
fore useful to reformulate the three basic ethical principles under proce-
dural maxims. Particular maxims often involve the application of one of
the three ethical principles in a restricted context. They often as well
combine more than one principle under a single procedural maxim because one
human activity ( e.g. , gaining free and informed consent) can serve more
than one purpose.
The basic procedural issues can be clustered under four headings: (1)
obtaining the free and informed consent of the human subject involved in
the research under consideration; (2) obtaining a proxy consent from indivi-
duals unable to consent in order to protect the best interests of those sub-
jects; (3) avoiding coercion which would unduly direct by threat the consent
of a human subject ( i.e. , one can be said to have chosen in a free-- e.g. ,
while not drugged--and informed fashion to give one's money to a mugger under
threat in the sense that one is a free agent and may be informed clearly of
the consequences of not handing over one's money, though one cannot be said
to have done it voluntarily). There is, in short, a distinction to be drawn
between the freedom of moral agents versus the liberty of uncoerced agents
(compare the contrast between the freedom of a normal adult versus its ab-
sence in a three-year-old child, and the contrast between the liberty of
choosing without coercion versus under duress). Thus, infants cannot give
free consent, normal adults can give free consent, though the consent of
8-13
some normal adults, perhaps prisoners, may not be voluntary; (4) one ought
to weigh the benefits versus the risks involved in the use of human subjects
in research in order to be assured that the costs of such research will not
outweigh its possible benefits. This procedural principle is a nearly per-
fect reiteration of the moral principle (C) above to maximize the benefits
accruable to society from research involving human subjects. It is the most
clearly teleologically-oriented procedural principle.
Because the four procedural principles outlined above reflect divergent
moral considerations, it is worth laying them out schematically to show the
diversity of issues at stake:
A. One ought to acquire the free and informed consent of
human subjects involved in biomedical or behavioral
research. Such a maxim can be seen to be derived from
at least three origins:
a. Respect for the person-subjects as moral agents
(ethical principle A).
b. Interest in encouraging autonomy as a general
social value (ethical principle C).
c. Concern to ensure that the best interests of the
subject can be secured (ethical principle B).
B. When true consent cannot be acquired because the subject is
not a moral agent ( e.g. , is an infant), proxy consent should
be obtained:
a. To ensure that the best interests of human subjects
will be considered (ethical principle B) within the
8-14
bounds of (i) the general duty of human subjects to
assist their society in recompense for benefits re-
ceived or likely to be received (a type of covenant
obligation that would be derived from respect of
the other members of society in principle A); (ii)
to preserve the general social interest in autonomy
by rehearsing a ritual of free and informed consent
on behalf of the human subject to be involved in re-
search, although that subject cannot consent on his/
her own behalf (ethical principle C); (iii) a con-
cern to preserve care and fellow feeling towards
human subjects, even those subjects not capable of
acting freely, because of our interest in those
sentiments (ethical principle C).
b. To establish a means to ensure that human subjects
who will become free agents will not have their
existence and abilities compromised before they can
act freely concerning their own lives ( i.e. , chil-
dren who will be moral agents and suffer the re-
27
suits of past care--ethical principle A).
One should avoid coercion of persons who are considering
volunteering as subjects in research, in order:
a. To preserve respect for persons (ethical principle A)
b. To increase autonomy in society, a goal of general
interest (ethical principle C).
8-15
D. One should weigh benefits versus risks in order:
a. To ensure that the benefits to subjects involved in
therapeutic experimentation outweigh the risks to
those subjects (ethical principle B). This can be
a paternalistic concern which would foreclose a
patient's ability to pursue an experimental form of
therapy which a disinterested observer might not
find justified considering the unlikelihood of suc-
cess and the amount of discomfort, or it can be a
concern to maximize utility.
b. To ensure that the benefits to society justify
the risk to the individual involved in non-thera-
peutic experimentation (ethical principle C). This
may also involve paternalistic considerations (and
thus ethical principle B)--a volunteer may wish to
risk his or her welfare by agreeing to participate
in an interesting and risky experiment with only
remote chances of revealing significant information,
and which would not in the eyes of a disinterested
observer be a justified risk.
As indicated by the points outlined above under each of the four proce-
dural principles, the considerations are diverse and turn on issues from the
respect of the freedom of persons to considerations with regard to protecting
or advancing certain goods and values.
8-16
III. OPINIONS IN THE CURRENT LITERATURE CONCERNING BASIC ETHICAL PRINCIPLES
WHICH SHOULD GUIDE THE USE OF HUMAN SUBJECTS IN BIOMEDICAL AND BEHAVIORAL
RESEARCH
The literature concerning the use of human subjects in experimentation
is immense, but relatively little of it bears on the question of basic ethi-
cal principles. What does, does so somewhat unsystematically. As a conse-
quence, what I will provide here is a typology of the various positions ar-
rayed according to the schemata of the ethical principles and procedural
maxims outlined above. In doing this, a few authors are presented in order
to cluster the issues around particular viewpoints concerning the nature of
the basic ethical principles involved in the use of human subjects. The
authors and materials have been chosen in order to display the range of
ideas with some clarity, rather than to give an inventory of the authors
writing on this subject.
A. The Role of Consent by the Human Subject
The concern for informed consent by the human subject involved in re-
search is clearly drawn from a sense of responsibility to respect persons
as moral agents by never treating them as means merely, but rather as en-
tities who have a right to their own self-determination. A very thorough
examination of the moral significance of consent is provided by Charles
Fried in Medical Experimentation . He begins from the legal intentions
that only consent which is free and informed "justifies what the law calls
OQ
'intermeddling' with a person's body." Out of this moral position in
the fabric of the law, Fried argues to four moral rubrics which focus on
medical procedures, including experimentation with human subjects: (1)
8-17
lucidity, (2) autonomy, (3) fidelity, and (4) humanity. By lucidity, Charles
Fried means that "the patient [but for that matter the subject in research
as well] has a right to know all relevant details about the situation he
29
finds himself in. This is an accent on knowledge sufficient to rational
action. By autonomy, Fried means the "liberty to dispose of one's self,
that is, of one's person, one's body, mind and capacities according to a
30
plan and a conception fully chosen for one's self." In his appeals to
fidelity, Fried wishes to exclude lying or deceit, and under the rubric
"humanity" he wishes to hold that "a person has a right to have his full
human particularity taken into account by those who do enter into'rela-
31
tions with him"
In all of these, Fried appears to be developing basic deontological
considerations in the context of research involving human subjects. His
position is very close to the Kantian categorial imperative that persons
"be treated at all times also as ends-in-themselves [ i.e. , as free agents],
32
never merely as means." This is not to say that Fried examines the issues
of individual rights to the neglect of social duties. He is very clear that
one of the central concerns with the morality of experimentation is the ex-
tent to which the bodies of persons "are at the disposal of the human groups
33
of which they are a part." He suggests in fact that the right to receive
medical care may be bound to a duty in at least some circumstances to con-
34
tribute as a subject to biomedical research. His accent, though, is upon
the rights of individuals to their own integrity, including the integrity
of their bodies and the right to be treated with honesty.
The value of informed consent, along with its function as a moral impera-
tive for treating persons with respect, is outlined in a useful but different
fashion by Alexander Capron. He lists the functions of consent as (1) to
promote individual autonomy; (2) to protect the patient-subject status as
a human being; (3) to avoid fraud and duress; (4) to encourage self-scrutiny
by the physician-investigator; (5) to foster rational decision-making; and
35
(6) to involve the public. His first consideration, to promote individual
autonomy, is the most clearly deontological , focusing on the duty to treat
human subjects with respect. And the second point, to protect the patient-
subject status as a human being, appears to be a further development of this
first point, while the third point, to avoid fraud and duress, involves for
Capron, sketching out procedures to protect the subject's freedom from the
zeal of the investigator. Thus, in the first three cases, Capron emphasizes
respect for the human subject as a free agent and involves a reiteration of
Fried's points with respect to lucidity, autonomy, and fidelity.
His fourth point, to encourage self-scrutiny by the physician-investiga-
tor, is a procedural consideration intended to provide a useful inventory of
the risks involved in the research, as well as to promote candor by the in-
vestigator to aid patients to distinguish between those elements of an in-
tervention which are therapeutically-oriented and those which are bent on
research alone. The fifth point, to foster rational decision-making, is a
further iteration of the fourth point with the accent falling on making the
human subject a better partner in the research in which he or she partici-
pates. Finally, the sixth point, making the issues involved in the free and
informed consent explicit so that the public can make better judgments con-
8-19
cerning the probity of particular research, involves both a general interest
in promoting autonomy as a general social value (an aspect of points four
and five), as well as other general social interests, including social con-
trol of the kind and quality of research.
To borrow a phrase made popular by Paul Ramsey, analyses such as those
by Fried and Capron concern how to treat the human subject in research as
a person, considering both to the duty to respect human subjects as free
moral agents, and the concern to maximize a general social value, that of
autonomy. One should notice, though, that these are two quite different
concerns. The first bears on maintaining the relationship between investi-
gator and subject as a moral one, as a partnership or a covenant between
subject and investigator. As Ramsey puts it, "consent expresses or esta-
blishes this relationship [ i.e. , between subject and investigator as per-
sons], and the requirement of consent sustains it. Fidelity is the bond
between consenting man and consenting man in these procedures." The
second is a concern to promote autonomy as a value, apart from any consi-
deration of autonomy as a basic right. Consider Professor Hans Jonas' re-
mark that "the individual's interest in his own inviolability is itself of
public interest such that is publicly condoned violation, irrespective of
numbers, violates the interest of all. In that case, its protection in
each instance would be a paramount interest, and the comparison of numbers
37
would not avail." There is, as well, a third consideration, that allow-
ing the individual subject free and informed consent increases the likeli-
hood that the choice will be in the interest of that subject, on the pre-
supposition that that subject is the best judge of his or her own best
8-20
interests. With respect to basic ethical principles, this consideration is
often reducible to either ethical principle A or ethical principle C, namely,
respect for persons or maximizing the general good in society through indi-
vidual initiative. Also, informed consent may derive from enlightened self
interest, a concern to allow others to choose in their best interests so that
we may similarly have that prerogative. Further, some have argued that in-
formed consent is justified in part because of the feeling of well-being it
gives to the volunteer. "It appears to this writer [John Fletcher] that the
conduct of the consent situation is decisive for the patient's or volunteer's
sense of being respected as a person, especially when the request is for a
OO
procedure which is non-beneficial."
In contrast to the arguments concerning the necessity of free and in-
formed consent, there is as well a set of arguments alleging the impossibi-
lity of free and informed consent: "The subject is ordinarily not qualified
to evaluate the true risks and expected benefits of any experimental drug
or procedure. In addition, an investigator who is eager to confirm some
hypothesis might, in informing the subject, minimize, either consciously
or unconsciously, experimental risks and uncertainties. Indeed, the investi-
39
gator may not know all the risks. ... " " Such arguments cannot be taken
seriously without eroding the \/ery possibility of recognizing free choice
in society generally. Persons, though never acting in any case as perfectly
free and rational agents, do so up to a point. The attempt to achieve in-
formed consent provides a criterion for moral action in the use of human
subjects. It provides one with a moral maxim: one should provide all the
facts material to a subject's decision (including the extent to which facts
8-21
or good information does not exist), and give that subject an opportunity
to acquire further facts to the point at which that subject is satisfied
with his or her understanding of the research at hand. And, again, the
absence of good information concerning some possible outcome of the experi-
ment can also be conveyed. (Laymen come into situations of incomplete or
minimal information more often than scientists assume—consider visits to
auto mechanics, or farmers judging next year's crop.)
To this must be added that third parties should ensure that the im-
portant risks and issues at stake are brought to the subject's attention--
a function at present served by DHEW mandated committees. One cannot try
(nor should one) to force subjects who can be rational free agents to use
that rationality and freedom to its fullest. Informed consent means that
the investigator should not circumscribe the free and rational character
of the subject's choice, but rather support that character as far as pos-
sible. The concept of informed consent thus functions to guarantee that
researchers will not act in disregard of their subjects as persons. It can-
not, however, function to guarantee that subjects will be ideal moral agents
who will act on their choices in a fully free and informed fashion. It is
only that the deficiency in their choice should not be attributable to the
investigator.
Informed consent as a means of preserving the interest of persons against
the interests of researchers in using them functions best through the pro-
tection of formal review committees. One should not conclude with Dr. Ingle-
finger that "the process of obtaining 'informed consent' ... is no more than
8-22
40
an elaborate ritual ..." because investigators can usually extract consent.
Rather, one should conclude that better public education and public safeguards
should be provided in order to support the free and informed consent of sub-
jects, and to avoid coercion in the obtaining of such consent.
However, a special set of experiments do require some level of decep-
tion and, therefore, some restriction of free and informed consent in order
to make them possible. These usually include randomized clinical trials of
drugs in which neither the investigator nor the patient knows whether the
patient is receiving a standard drug and/or a second drug and/or placebo used
in comparison. From the argument above, it would follow that one would owe
to the patient (1) clear knowledge that he or she was the subject of a ran-
domized clinical trial ( i.e. , might be receiving, for example, a placebo or
a less well-known drug); (2) assurance that there was good evidence that
being subjected to this risk of receiving a non-standard drug or placebo
would not be a substantial risk to the patient (a proper paternalistic con-
cern); and that (3) provision was made to evaluate the difference between
the therapies so that should one therapy show itself to have a particular
advantage or a particular risk, further trials would be discontinued. One
needs, in short, to meet two central moral issues involved in randomized
clinical trials for, to quote Charles Fried, "The protocol may require an
abdication of professional judgment such that the patient's therapy is
determined not by the needs of his particular case, but by the demands of
the experimental design," and "The practice of RCT's [randomized clinical
41
trials] often, though not necessarily, involves deceit." Given a full
disclosure of the facts of the trial, and given a careful assessment that
8-23
a random clinical trial is merited because of the uncertainty with respect
to the merit of the treatment or treatments currently available, there are
not insuperable problems. Individuals can be given sufficient information
so that they can decide to assume a minor risk for the acquisition of medi-
cal knowledge. One is thus brought back to the basic ethical principle A,
to treat the subject with respect, and basic principle B, to protect his or
her best interests. Though consent in the contexts of research rarely pos-
sesses the clarity to justify the maxim "Volenti non fit iniuria," it is
rarely so muddled or invincibly in the dark that a proper presentation of
information will not allow a morally adequate choice.
There is though, a genre of experiments which depends necessarily upon
an element of deception—usually experiments in psychology or social psycho-
logy. They are often innocuous studies in which the only violence done to
the subject is being told a lie concerning the actual aim or goal of the
research. Others have included substantial psychological, as well as
moral, risks. Some of these experiments have indeed provided varying degrees
of informed consent. For example, Professor Philip G. Zimbardo, in reflecting
on critiques of his Stanford prison experiment, states that "the 'informed
consent' statement, signed by every participant, specified there would be an
43
invasion of privacy, loss of some civil rights and harassment." " Experi-
ments such as Stanley Milgram's have involved subjects in circumstances with
both an element of deception and of risk. Subjects were asked, for example
to apply electroshock to a second person, in a controlled social setting
(under the ruse of participating in a scientific experiment). One should
note that there is not only a psychological risk, but a moral one as well--
8-24
the subject may, in such experiments, be caused to perform an action which
he or she would hold to be immoral. That is, a psychological experiment
can function as a form of seduction with not only psychological but ethical
consequences for the subject involved. Experiments should not be designed
which would entice persons to violate moral maxims, because to do so is to
fail to treat the subject as a moral agent, as worthy of respect. One
fails to acknowledge the subject's moral integrity.
It is worth noting that The Ethical Principles in the Conduct of Research
with Human Participants of the American Psychological Association provides
for deception if the researcher shoulders the moral responsibility of that
deception. For example, "Failure to make full disclosure gives added empha-
sis to the investigator's responsibility to protect the welfare and dignity
of the research participant. ... The decision to limit this freedom [ i.e. ,
to decline to participate in research or to discontinue participation at
any time] increases the investigator's responsibility to protect the parti-
46
cipant's dignity and welfare."
There are at least three issues at stake with respect to free and informed
consent in such research: (1) whether active deception is permissible—whether
researchers have a right to lie in the service of science; (2) whether re-
searchers have a duty to disclose the nature of a psychological experiment,
if an experiment has the possibility of attendant risks, even in the absence
of active deception; (3) whether in the absence of active deception, or with
a degree of deception not materially different from that of everyday social
interactions ( e.g. , polite evasion of the truth), an investigator has a duty
8-25
to disclose the nature of an experiment when no risks to that subject, in-
cluding invasion of privacy, are involved.
Respect for persons, it would appear, precludes any active deception
in the performance of an experiment. Persons as such have a right not to
be used as means merely. But, depending on the nature of the experiment,
it may at times be possible to inform subjects that they will be set at
risk of some deception, and thus avoid using them as means merely, while
still avoiding disclosing the particular variables under study. ' In some
cases, then, when the risk is minimal, general consent can be reasonably
given to a risk of some deception as part of an experiment protocol. The
crucial moral issue is that there not be greater deception, nor greater
risk of psychological or ethical harm, than the subject would have antici-
pated in his or her consent, and that that risk, in any event, not be sub-
stantial. Beyond that, deception which one has accepted should be consi-
dered as morally harmless, as is the systematic deception which is part of
a well -played game of poker.
There should be, as well, a recognition of the right of the subjects not
only to be free of active unaccepted deception, but there should be disclo-
sure of relevant information when an experiment sets that subject at risk,
even in the absence of deception. It is incompatible with respect for per-
sons as free moral agents to set them at a risk without their consent, for
such action would be a paradigm case of using another as a means merely, as
an object rather than as another person. The only circumstances which
might be exceptions to this rule would be those in which the research was
8-26
to the direct benefit of the recipient and the experimental nature of the
procedure was not relevant to the patient's choice. One might be able to
conceive of cases in which an experimental therapy is the treatment of choice
for a patient with a serious disease, and the patient is in circumstances
such that a full disclosure of the nature of the therapy would worsen the
48
patient's prognosis. Those cases are probably very rare. Moreover, phy-
sician-experimenters would be morally justified in withholding such know-
ledge only when they were sure that the patient would indeed want to pursue
such therapy. After all, patients have a moral right, and an increasingly
recognized legal right, to refuse therapy, including life-saving therapy.
Such decisions would have to be made with every reason to believe that one
would not be acting against the choice of the patient.
Finally, there is a class of experiments which involves neither risk
nor intrusion, but simply observation. The American Psychological Associa-
tion gives the example of observing transportation patterns or other forms
of natural public behavior which are, in any event, open to everyone's cri-
tical observation. Though the researcher does not have any rights beyond
those of the normal citizen to lie and deceive, the researcher does not
have a special obligation to make known his or her projects to others if
those projects are non-invasive and carry no risk to the subjects involved,
apart from the general risks of everyday life ( e.g. , being observed while
doing what people usually do on the public streets). Such research could
reasonably include that element of deception which is part of everyday life
--the researcher may hold a job as a part of a research endeavor in order
to better observe worker interaction, without divulging to the workers in
8-27
50
advance his true identity. That is, one can presume that people consent
to a minimal level of disingenuous behavior as an element of public life
( e.g. , no one tells the host that his choice of wine is poor). The line
between harmless passive deception of others which is part of usual life
situations ( i.e. , hiding elements of one's true thoughts, etc.) versus
active deception of an active fashion that departs from this level, is,
as are all lines in real life, difficult to draw.
All these concerns coalesce around the three points raised under the
procedural maxim of obtaining free and informed consent of human subjects
participating in biomedical and behavioral research. Respect for human sub-
jects as persons requires gaining their free and informed consent in order .
to use them, not as mere objects, but as collaborating persons in an experi-
ment. The exceptions are those where in fact no use if made of another--
research involving only the observation of everyday deportment (where con-
sent to observe is implied in the very act of entering a publicly observable
sector of the world). And, again, the other roles of consent must be empha-
sized: to encourage autonomy as a general social value, and to ensure that
the best interests of the subject are secured.
B. The Role of Proxy Consent
One of the most contested issues concerning experimentation involving
human subjects is the role of proxy consent. It is surely a strange notion-
something like having a friend drink a glass of water for you when you are
thirsty. If consent functions as a way of respecting the freedom of indivi-
duals by gaining their leave in order to use them in research, then proxy
8-28
consent rarely makes sense (it would in cases where individuals have indeed
enacted a power of attorney in order to convey to others the right to make
decisions on their behalf while they are incapacitated).
A number of individuals have on the basis of respect for persons argued
against the concept of proxy consent. Paul Ramsey, for example, would allow
proxy consent only in cases where such consent is made on behalf of the
manifest good of the individual concerned — for example, a parent consenting
to therapeutic experimentation, with the prime intention of choosing the
form of therapy most likely to be of use to the child. To quote Ramsey,
"From consent as a canon of loyalty in medical practice it follows that chil-
dren, who cannot give a mature and informed consent, or adult incompetents,
should not be made the subjects of medical experimentation unless, other
remedies having failed to relieve their grave illness, it is reasonable to
believe that the administration of a drug as yet untested or insufficiently
tested on human beings, or the performance of an untried operation, may
further the patient's own recovery ... consent-requirement means: 'Never
submit children to medical investigation not related to their own treatment,
51
except in face of epidemic conditions endangering also each individual child.'"
Ramsey's objection is based on the contention that to make a child subject to
an experiment involving any risk or discomfort is to use that child as a means
merely, rather than acknowledging it with the respect due to persons.
In short, Ramsey's argument is that one should treat all humans the same,
and, since one treats adult humans with respect due to free agents, one
should do the same with respect to infants. He puts this very strongly,
"[children] can be harmfully used, or they can simply be used with no harm.
8-29
Both the degradation of the body's fortress and being treated as a means
only are human violations." 5 This position would exclude children from
any non-therapeutic research, except passive observation involving no touch-
ing or other intrusion. Ramsey's position evidently is meant to apply to
all humans who might be incapacitated, and precluded from giving free and
informed consent.
The preponderance of those engaged in biomedical and behavioral research
appear to accept more readily using infants and other incompetents. Though
the Nuremberg Code omits mention of the use of incompetents, provision for
this is explicitly make in the Declaration of Helsinki ("If he is legally
incompetent the consent of the legal guardian should be obtained"). Curran
and Beecher must be counted in this group. They have argued for allowing re-
search involving immature children (those under fourteen years of age) when
there is no discernible risk. "Not to allow such studies would greatly hamper
important nutritional, psychological, and educational studies in children,
54
as well as studies of inborn errors of metabolism and genetic defects."
That is, there are special issues raised concerning the physiological and psy-
chological responses of children and the mentally ill that can only be answered
by research involving these populations.
Moreover, treatment tried for the first time in special populations
( e.g. , children) constitutes an experiment. As Leon Eisenberg points out
"There is a non-trivial risk whenever a drug is given for the first time to
a child. Further, the more potent the drug in treating the condition at
55
which it is directed, the greater the risk of undesired side-effects."
8-30
Such experimentation is particularly troublesome when, as in the case of the
rubella vaccine, its prime intention was not to prevent rubella in the chil-
dren vaccinated, but in the fetuses which some of those children might bear.
To encompass experimentation such as the trial of the rubella vaccine, one
would not only have to allow intrusive experimentation on incompetent popu-
lations redounding the benefit of that population ( e.g. , aiding in the
development of treatment of childhood diseases or mental illnesses), but
would also have to include research that cannot be done in other popula-
tions ( e.g. , use of rubella vaccine in pediatric age populations would al-
ways be, at the time of its first introduction, experimental --and insofar
as the major goal would be to prevent fetal defects, it would to that ex-
tent not be for the benefit of the population of children. In order to
have a rubric to encompass all such experimentation on children and other
incompetents, one would need a rule somewhat similar to that forwarded by
the AMA, that "minors or mentally incompetent persons may be used as sub-
jects only if: (1) the nature of the investigation is such that mentally
competent adults would not be suitable subjects; (2) Consent, in writing,
is given by a legally authorized representative of the subject under cir-
cumstances in which an informed and prudent adult would reasonably be ex-
pected to volunteer himself or his child as a subject. ' One would, in
short, have to have a rule which sanctioned experimentation on incompetents,
when that experimentation was non-therapeutic and involved very minimal risks
and discomforts.
There have been several attempts to resolve the problem of the use in
research of incompetents and, in particular, children. Suggestions have been
8-31
made of a procedural sort, namely, that one should move progressively into
younger age groups, thus avoiding any very unexpected reactions in the intro-
58
duction of new drugs. On the other hand, Richard McCormick has attempted
to give a solution in principle. He wishes to justify those experiments
involving children "that are scientifically well-designed (and therefore
offer hope of genuine benefit), that cannot succeed unless children are
used ... [and] that contain no discernible risk or undue discomfort for
CO
the child. 3 McCormick's position thus contrasts with Paul Ramsey's, for
McCormick would allow subjecting children to discomfort. McCormick justifies
such use ( i.e. , subjection to discomfort) of the child on the basis of what
the child ought to wish to do. "To share in the general effort and burden
of health maintenance and disease control is part of our flourishing and
growth as humans. To the extent that it is good for all of us to share
this burden, we all ought to do so. And to the extent that we ought to do
so, it is a reasonable construction or presumption of our wishes to say
that we would do so ... sharing in the common burden of progress in medi-
cine constitutes an individual good for us all up to a point . " It is in
this fashion that McCormick justifies extensive experimentation and research
upon children which is morally foreclosed according to Ramsey's account.
Both Paul Ramsey and Richard McCormick frame their arguments as if one
had to consider the will or wishes of incompetents. Yet, it is precisive
because incompetents have no will, in the sense of a moral will, that they
are incompetent, they cannot choose. There is, thus, no contradiction
between the AMA's statement that minors and mentally incompetent persons
can be used in research on the basis of proxy consent, and the assertion
8-32
that "no person may be used as a subject against his will." One does not
violate the will of a baby as one brings it kicking and screaming to a
harmless pin prick as its sole contribution to a hematologist's study of
small infants. There is no one's will or freedom to violate. Infants,
though often willful, have no free will, and are not the object of respect
in the sense that adults are. Thus, one respects the right of a Werner
CO
Forsmann to catheterize his own heart at unknown and, perhaps, considera-
ble risk, while one does not accept the free consent of a normal three-
year-old child to similarly volunteer itself. Proxy consent does not exist
to respect the child as a moral agent, but rather to safeguard its best
interests.
If proxy consent is not seen as a safeguard against the violation of
an infant's wishes (on the grounds that it is not a free agent as normal
adults are), but rather as a means to protect its best interests, it follows
that that experimentation that does not involve an increase of risk over the
ambiance or any significant discomfort, should be allowed. One is not vio-
lating anyone's moral integrity. The point is rather to preserve the physi-
cal and psychological integrity of the child. Experiments that do not set
the physical or psychological integrity of the child, or other incompetents,
at risk should be prima facie allowable.
McCormick's point can then be reformulated, not in terms of what the
child ought to wish, but in terms of what minor increase in risks should
reasonably be borne by us all as members of a society that has chosen to
pursue certain goals, including the improvement of health care without
8-33
CO
appearing severe or undue. This final issue is too complex for the scope
of this paper. The most that can be indicated is that nearly all societies
recognize a certain minimal level of risk as a social obligation to which
one may be said to have implicitly consented by ones presence. Of course
if they were competent, the children might protest, even to the point of
demanding exit from the society. But insofar as they are not, they are not
bearers of freedom, and need not be respected. They are though bearers of
interests, and therefore to be cared for. Thus, proxy consent functions not
as a way of respecting incompetents as moral agents, but as a way of pro-
tecting their best interests. Therefore, experiments that would not act
against their interests, by exposing them to physical or social psychologi-
cal risks, greater than those in the usual ambiance, are prima facie proper.
Experiments involving very minimal risks (nearly that of the subject's am-
biance) and minor discomfort, should be justifiable in terms of an appeal
to the minimal duties that each of us owe to our society. There will, in
the second case, surely be no clear lines to be drawn. One would hope that
such lines will be drawn by prudent persons. But the absence of black and
white distinctions should not cause one to retreat into the darkness, rather
than to attempt to draw reasonable lines in the twilight.
C. Avoiding Coercion in the Context of Research
The issue of coercion in research involving human subjects usually in-
volves special populations such as students who may feel forced by their
teachers to volunteer, prisoners who may feel coerced to volunteer because
of their circumstances, or the poor who may be forced to volunteer because
of the situations in which they receive their health care. I will not deal
8-34
with the issues of these special populations, except to make this point: the
morality of the coercion brought to bear upon such populations cannot be as-
sessed except in terms of very basic judgments bearing on the morality of the
situations in which those populations live. It is thus not possible to de-
cide whether prisoners should usually be barred from participating in non-
therapeutic research, until one has decided what is due prisoners. One must
first answer certain basic questions regarding the theory of punishment.
Hans Jonas raises this issue in a very jarring fashion, "If we hold to some
idea of guilt, and to the supposition that our judicial system is not entirely
at faults, [prison inmates] may be held to stand in a special debt to society,
and their offer to serve—from whatever motive—may be accepted with a mini-
mum of qualms as a means of reparation." Apart from whatever judgment one
might have concerning Professor Jonas' statement, an adequate account of the
use of prisoners must include an account of the role or rationale of the
practice of punishment--and that lies beyond the scope of this paper.
In any event, action to prevent coercion in the procurement of consent
to research is motivated out of respect for the integrity of persons as well
as a concern to preserve autonomy as avalue in society. Both issues, for
example, appear in the Kaimowitz v. Doe case where the court held that "in-
voluntary confined patients cannot reason as equals with doctors and admin-
istrators over whether they should undergo surgery. They are not able to
voluntarily give informed consent because of the inherent inequality in their
position." One sees in such cases an attempt to discern the geography
of offers and threats in order to distinguish between proper inducements to
participate in research ( e.g. , satisfaction of contributing to science,
8-35
performing a socially approved act, making retribution for one's previous
crimes, etc.), and those inducements which the Nuremberg Code indicated
under the rubric of "duress, overreaching, or other ulterior form of COn-
straint or coercion." That distinction must be drawn on a case by case
basis in order to prevent both coercion and foreclosure of free choice out
of an inordinate fear that coercion is present. Coercion is an evil in
the context of research because it bears both against the freedom of the
subject as well as against the good of society. But an overzealous at-
tempt to avoid the possibility of coercion ( e.g. , by categorically forbid-
ding all non-therapeutic research with prisoners which does not bear
directly on the good of prison populations) may deprive individuals of that
very liberty one wished to protect. The issues here (apart from basic ques-
tions of the nature of punishment) are, for the most part, procedural rather
than questions bearing on basic ethical principles.
D. Sufficient Benefits to Justify the Risk
It is not vicious, but in fact virtuous to attempt to maximize the goods
available to us all. A utilitarian calculus set within the bounds of respect
for persons as free agents can be a noble way of living in a world defined
by scarce resources. One should not invest human resources and energies to
no avail. This virtue becomes a duty in the use of public funds. Such con-
siderations apply a fortiori to biomedical and behavioral research involving
human subjects.
It is under this rubric of cost/benefit analysis that conflicts between
individual versus societal rights have a natural place. They arise because
8-36
a society will be tempted to pursue experiments that promise considerable
benefit at the cost of damage only to a few members of society, perhaps indi-
viduals already in a socially disadvantaged position. Those who lean toward
the utilitarian viewpoint emphasize the benefits to mankind of human experi-
mentation and are sometimes willing to suggest that the benefits of such re-
search are so enormous as to constrain us to recognize a "duty" of subjects
to participate in it. At the very least, what the utilitarian sees is a
calculus of goods and values. As it was succinctly put by a contemporary
writer, "Human experimentation is required in order for medical progress
to be made. Such research will of necessity involve risks that can be mini-
mized, but not eliminated. ... The problem, therefore, boils down to a sober
weighing of costs and gains. ..."
Balancing the expected good to result from research against the risk
to the individual subject is, though, an ambiguous undertaking. It can
have at least three meanings: (1) it can simply indicate a cost/benefit
analysis for society in which one measures the good that would accrue to
society from that research against the loss to society as a result of any
damage to the individuals participating. That is a proper calculus, if
placed within the bounds of respect for human freedom. A society may thus
decide not to support research which on the balance would be costly to it,
even if there were individuals free and willing to volunteer for it; (2)
there is also the sense of balancing the expected benefit to the individual
from the research against the foreseeable risks to the individual --an ele-
ment of giving sufficient information for adequate consent to a subject
considering entering a therapeutic experiment; (3) finally, society may
8-37
have a basis for forbidding individuals from participating in research with
high potential risks and low potential yield. From paternalistic motives
to act in the best interests of the would-be subject on the assumption that
society, or some other organ of society, knows the best interests of that
individual better than the individual, him or herself. But the difficulty
with free agents, as indicated above, is that the defining characteristic,
upon which the moral significance of values hinges, is to act freely.
Therefore, it may be one thing to decide not to support such research
through public funds, and another thing to try to prevent it through any
other coercive methods. An element of a free society would seem to be that
one can either jump motorcycles over gorges to fame, or participate in reck-
less experiments as long as the rights of others are not materially affected.
This last point, thus, concerns balancing goods within the bounds of
respect for freedom. Since we both respect persons and have interest in
goods and values, we must allocate scarce resources as best we can to for-
ward our commonly embraced goals within the bounds of never treating other
persons as means merely. One may never treat others as means merely . But,
one may, within the bounds of moral probity, treat others as means. In-
formed consent of the other one is the difference between using another as
a means, and using another as a means merely. Which is to say, ethical
principles (B) and (C), bearing on the best interests of individual subjects
and the best interests of society, find their locus within the bounds of
respect for human subjects as free agents (principle A). In the end, what
distinguishes the moral problem of experimentation with human subjects
from the moral problems of experimentation with subjects of other species,
is that only persons can make absolute claims to respect.
8-38
IV. SUMMARY
Moral postures concerning research involving human subjects are diverse
in part because of the diversity of issues at stake in, for example, disputes
over conflicts of rights and duties, concerning values, and with regard to
duties to persons with interests in goods and values. I have presented
three ethical principles around which to gather these questions: (1) respect
for human subjects as free agents; (2) concern to foster the best interests
of human subjects; and (3) concern to maximize the benefits accruable to
society. I have indicated that these three abstract considerations arise
around four procedural foci for research involving humans: (1) the require-
ment of free and informed consent of competent human subjects; (2) the require-
ment of proxy consent from incompetent human subjects; (3) the avoidance of
coercion in the consent context; and (4) an interest in having research in-
volving humans redound to the general good of society. These are at best
guidelines, or outlines of the central ethical issues at stake. They must
in each case of research be applied with care and followed with prudence.
One can never have a means of simply deducing answers. Basic ethical prin-
ciples represent, rather, our best attempt to map out the terrain of rights
and values.
I am in debt to Edmund L. Erde, David Ost, Michele Malloy, and John Moskop
for their ideas and suggestions during the development of this paper. The
fact that they engaged in active discussion of the ideas in this paper does
not imply their agreement with them.
8-39
FOOTNOTES
1. These two questions can be distinguished as deontological versus
teleological moral considerations, that is, concerns with rights and
duties versus goals. As John Rawls succinctly puts it, "The two
main concepts of ethics are those of the right and the good ... a
deontological theory [is one] that either does not specify the good
independently from the right, or does not interpret the right as
maximizing the good. ... Now it seems that the simplest way of re-
lating them is taken by teleological theories: the good is defined
independently from the right, and then the right is defined as that
which maximizes the good." A Theory of Justice (Cambridge, Mass.,
1971), pp. 24, 30, 24.
2. Consider, for example, the attitudes of behaviorists, who interpret
behaviorist psychology not as a methodology but as a form of meta-
physics. See B. F. Skinner, Beyond Freedom and Dignity (New York,
1971).
3. What I am proposing here is a view of freedom and morality that
draws heavily upon Immanuel Kant, that freedom is a presupposition
for both claims to knowledge and morality. See Immanuel Kant,
Critique of Pure Reason , A 542=B 570 to A 558-B 586. Foundation
of Metaphysics and Morals ( Grundlegung der Metaphysik der sitten ) ,
in Kantswerke , Akademic Textausgabe (Berlin, 1968), Vol. 4, pp. 446-
447.
4. A reduction can be made of any position holding that persons are
not free in this fashion: one could disagree with the proposition
that persons are free only on the basis of presupposing that persons
are free, that is, by presupposing that one choose to hold one's
position ( i .e. , that persons are not free) on the basis of good
reasons.
5. The Nuremberg Code, in CIOMS Round Tables , ed. V. Fattorusso (Paris,
1967), p. 100.
6. It follows that the fact that infants and other incompetents cannot
be treated as persons in the same way that normal adults can, that
humans are persons in different ways. Normal adult humans are per-
sons in the strict sense of being actual moral agents, bearers of
rights and duties. Small children and other mental incompetents are
treated as if they were persons, moral agents, insofar as this is
possible. That is, they are treated as having moral rights, but not
moral duties ( e.g. , an infant can be said to have a right to life, but
not a duty to preserve the lives of others). Moreover, though infants
and other incompetents can be said to have rights to certain goods,
they do not have a right to be treated as free moral agents, which
8-40
they are not. This distinction is of importance with respect to
proxy consent, and Paul Ramsey's position which is discussed below.
7. In this regard, Kant offers the useful distinctions between the
dignity of persons and the values of things. A person may be
treated as having a particular value ( e.g. , a utility value), but
must also be acknowledged as having the dignity of a free agent.
A person has the right to be treated not merely as an object, but
also as self -determining—an entity who can be used only with his
or her consent. Metaphysik der sitten , in Akademic Textausgabe ,
Vol. 6, pp. 434-435.
8. In this sense a person has a right to be treated as a free agent
in human research, not because treating him or her otherwise would
erode the fabric of society, etc., but simply because that is
what is due that human subject.
9. Talk about rights and duties also occurs within teleological theories
of morality. In such contexts, it really stands for concern for some
good or value. In such a framework, the subject's right to free and
informed consent would express a judgment that in the absence of
such a practice members of society might be put in jeopardy, feel
ill at ease, etc. Such talk of rights and duties is a disguised way
of talking about or enjoining one to action upon goods and values.
Consider, for example, John Stuart Mill's treatment of the status
of rights in Utilitariansim . "To have a right, then, is, I conceive,
to have something which society ought to defend me in the possession
of. If the objector goes on to ask why it ought, I can give him no
other reason than general utility." Utilitarianism , ed. Oskar Piest
(Indianapolis, 1957), p. 66. "... all cases of justice are also
cases of expediency ..." (p. 79).
10. Note that here the interest in human autonomy is a teleological one,
a good the pursuit of which has general social value.
11. That is, giving such respect to persons is a conceptual condition,
part of what is required in talking about responsibility to persons
as free agents. It is not a material condition, as, for example,
regular adherence to a utilitarian rule is a condition for mainte-
nance of a moral practice.
12. It is in this sense that respect for freedom is not a value, but the
condition for the possibility of responsibility to persons as moral
agents. But treating persons as free does not preclude making evalua-
tions concerning persons in a utilitarian fashion, as long as such
evaluation does not preclude respect for those persons as moral
agents. E.g. , one might not give persons the opportunity to consent
to research of little benefit through precluding public funding of
such research.
8-41
13. Arguments that the conditions under which prisoners are held pre-
vents them from giving voluntary consent to any experimental proce-
dures are relevant here. See, for example, Kaimowitz and Doe v .
Department of Mental Health for the State of Michigan , Civil Action
No. 73-19434-AW (1973). The ruling distinguishes among three ele-
ments of adequate consent: competency, knowledge, and voluntariness.
14. The Nuremberg Code, p. 100.
15. Declaration of Helsinki, in CIOMS Round Tables , p. 103.
16. "Department of Health, Education and Welfare Rules and Regulations:
Protection of Human Subjects," Federal Register , 39 (May 30, 1974),
p. 18917, § 46,3c.
17. American Psychological Association, Inc., Ethical Principles in the
Conduct of Research with Human Participants (Washington, D.C. , 1972) ,
p. 1 f.
18. The Nuremberg Code, p. 100 f.
19. Declaration of Helsinki, p. 103.
20. Declaration of Helsinki, p. 103,
21. DHEW Rules and Regulations, p. 18914.
22. American Psychological Association, Inc., p. 2.
23. The Nuremberg Code, p. 100.
24. Declaration of Helsinki, p. 102.
25. DHEW Rules and Regulations, p. 18917, § 46.2. As the reader will note,
this quote from the DHEW regulations is nearly the same in wording as
the one under (B) above. Both quotes contain the accent on the best
interests of the subject, as well as on the benefits accruable to society.
26. American Psychological Association, Inc., p. 2.
27. The duty to treat infants with care is thus greater than the duty to
so treat incompetents such as the permanently and severely senile,
because children will become free moral agents whose integrity as
such will have been compromised by those who failed to give them
due care. It is to persons in the strict sense, as free moral agents,
that we owe the basic duty of moral respect.
28. Charles Fried, Medical Experimentation (New York, 1974), p. 19.
29. Fried, p. 101.
8-42
30. Fried, p. 102.
31. Fried, p. 103.
32. Kant, Grundlegung der Metaphysik der Sitten , in Kantswerke ,
Akademic Textausqabe (Berlin, 1968), Vol. 4, p. 429.
33. Fried, p. 4.
34. Fried, p. 157.
35. Alexander M. Capron, "Informed Consent in Catastrophic Disease^Re-
search and Treatment," University of Pennsylvania Law Review , 123
(December, 1974), 364-376.
36. Paul Ramsey, The Patient as Person (New Haven, 1970), p. 5.
37. Hans Jonas, "Philosophical Reflections on Experimenting with Human
Subjects," Daedalus , 98 (Spring, 1969), 222.
38. John Fletcher, "Human Experimentation: Ethics in the Consent
Situation," Law and Contemporary Problems , 32 (Autumn, 1967), 632.
39. Robert D. Mulford, "Experimentation on Human Beings," Stanford
Law Review , 20 (November, 1967), 106.
40. F. J. Inglefinger, "Informed (but Uneducated) Consent," New England
Journal of Medicine , 287 (August 31, 1972), 466.
41. Charles Fried, Medical Experimentation , p. 148.
42. For example, see Jerome H. Resnick and Thomas Schwartz's discussion
of an experiment in which a control group was told that the experi-
ment's goal was to determine how college students form sentences,
though the goal was to determine the effect of verbal reinforcement
of certain words used by the subjects while doing the experiment.
One group of these individuals was treated simply as controls ( i.e. ,
no further information was supplied); another group was told that they
were indeed being employed in such an experiment. The real experiment
was to find out whether telling the truth concerning the experiment
would alter the results of the reinforcement of certain usages of lan-
guage. It did. Cf. "Ethical Standards as an Independent Variable in
Psychological Research," American Psychologist , 28 (1973), 134-139.
43. Philip G. Zimbardo, "On the Ethics of Intervention in Human Psycholo-
gical Research: With Special Reference to the Stanford Prison Exper-
iment," Cognition , 2 (1973), 243-256.
44. Milgram's subjects were paid $4.50 for participating '"no matter what
happened. 1 Yet, it is unlikely that they consented to being stressed
8-43
and some were apparently below the age of legal consent. It can
also be argued that the researcher's motivation was altruistic,
to contribute to recorded scientific knowledge. And valuable
behavioral data were recorded—over 50% of the subjects continued
to administer what they thought was full, actual shock, even after
the confederate expressed great pain." Robert I. Gordon, "Mental
Distress in Psychological Research," Baylor Law Review , 21 (1969),
526.
45. Stanley Milgram, "Some Conditions of Obedience and Disobedience to
Authority," Human Relations , 18 (1965), 57-75. Reprinted in Jay
Katz, Experimentation with Human Beings (Hartford, 1972), 358-
365. "[In one variation] the victim received a shock only when
his hand rested on a shock plate. At the 150-volt level, the vic-
tim again demanded to be let free and, in this condition, refused
to place his hand on the shock plate. The experimenter ordered
the naive subject to force the victim's hand onto the plate."
Katz, p. 361.
46. Ethical Principles in the Conduct of Research with Human Partici -
pants , § 3, 5, pp. 1 , 2.
47. One might ask whether classical studies such as those by Robert
Rosenthal and Lenore Jacobsen (which involved deception of teachers
as well as parents and students with respect to the intellectual
abilities of those students in order to test how expectations in-
fluence performance) could have been performed had the teachers
and parents been given a blanket warning that some sort of study
was underway. Also, a crucial question at issue is whether there
was indeed any risk to the students involved. That is, even if the
teachers and parents would have been willing to agree to partici-
pate in an experiment of an undisclosed nature, the crucial issue
would be whether there was minimal risk so that such a venture
would have been justified. See Pygmalion in The Classroom (New
York, 1968).
48. The Drug Amendment's Act of 1962 (Kefauver-Harris Bill) removes
the need for consent to the use of investigational drug when phy-
sicians hold "in their professional judgment [that such disclosure
would be] contrary to the best interests of such human beings."
Section 505(i).
49. American Psychological Association, Inc., p. 30f.
50. American Psychological Association, Inc., p. 32.
51. Ramsey, The Patient as Person , p. 11 f.
52. Paul Ramsey, "The Ethics of a Cottage Industry in an Age of Com-
munity and Research Medicine," New England Journal of Medicine ,
284 (April 1, 1971), 704.
8-44
53. Declaration of Helsinki, III, 3 a, p. 103.
54. William J. Curran and Henry K. Beecher, "Experimentation in Chil-
dren," Journal of the American Medical Association , 210 (October 6,
1969), 83.
55. National Academy of Sciences, Experiments and Research with Humans :
Values in Conflict (Washington, D.C., 1975), p. 97.
56. Someone might retort that such experimentation did redound to the
benefit of the population of children, namely, those fetuses who
would not be born deformed due to rubella, thanks to the immuniza-
tions. But fetuses are not quite the same as children, for among
other things, women exposed to rubella have abortion as a means of
preventing the birth of children who might be thus deformed. In
a real sense, the control of rubella exists for the good of parents.
57. American Medical Association, Opinions and Reports of the Judicial
Council (Chicago, Illinois, 1971 ) , p. 12.
58. Alexander Capron, "Legal Considerations Affecting Clinical Pharma-
cological Studies in Children," Clinical Research , 21 (February,
1973), 141-150.
59. Richard A. McCormick, "Proxy Consent in the Experimental Situation,"
Perspectives in Biology and Medicine , 18 (August, 1974), 14.
60. McCormick, p. 12 f.
61. Judicial Council Opinions and Reports , p. 12. My arguments have in-
volved children not capable of free choice--not older children who
are and constitute a different and more complex problem. Children
and other incompetents should be involved in all such decisions inso-
far as they are able.
62. Nobel Lectures, Physiology or Medicine , 1942-1962 (Amsterdam: Elsevier
Publishing Company, 1964), p. 511.
63. An important critique of this position is given by Professor Hans
Jonas, "Philosophical Reflections on Experimenting with Human Sub-
jects," esp. p. 229 f.
64. Jonas, p. 246 n.
65. Kaimowitz v. Doe , p. 29.
66. Nuremberg Code, § 1, p. 100.
67. Louis Lasagna, "Special Subjects in Human Experimentation," Daedulus ,
98 (Spring, 1969), 461.
8-45
9
MEDICAL ETHICS AND THE ARCHITECTURE OF
CLINICAL RESEARCH
Alvan R. Feinstein, M.D.
and
Jeffrey L. Lichtenstein, M.D.
MEDICAL ETHICS AND THE ARCHITECTURE OF CLINICAL RESEARCH
Alvan R. Feinstein, M.D.*
and
Jeffrey L. Lichtenstein, M.D.**
The current concerns over ethical issues in medical research arise at a
time of unprecedented achievements in the research. Tuberculosis, poliomye-
litis and other dread infectious diseases have become controlled or elimi-
nated. Cardiac surgery, prosthetic joints, and renal transplantation or
dialysis have transformed people's lives. Psychotropic drugs have dramati-
cally improved the outlook for mental disturbance. These and many other
advances in medical progress have made the continued conquest of human ail-
ments become an expectation, not just a hope.
The clinical investigation that led to this progress, however, has
recently come under intensive ethical scrutiny. Investigators have been
castigated for alleged laxity in the conduct of the research, for occasional
failure to attend to the basic rights of subjects, and for apparent exploita-
tion of the poor, the ignorant, or the incarcerated. Unethical medical re-
search has become a topic for deliberation by symposia of philosophers,
attorneys and theologians; for prevention by institutional review committees;
and for rejection by medical granting agencies and editors.
That medical research is necessary for medical progress has been uni-
* Professor of Medicine and Epidemiology, and Director, Johnson Clinical
Scholar Program, Yale University.
** Johnson Clinical Scholar and Postdoctoral Fellow in Medicine, Yale
University
9-1
versally accepted. The main sources of current dispute have been not the .,
need for research, but the kinds of research that are ethically permissible
and the kinds of principles that distinguish ethical from unethical research.
During the examination of these issues, the topic of human experimentation
has often been regarded as a unitary activity, whose problems might be amena-
ble to solution with such simple, single procedures as "informed consent."
Our purpose in this paper is to describe the complex intellectual
architecture'^' of projects in medical research, to indicate the diverse
ethical problems' ^raised by component elements of the architecture, and
to suggest that many ethical issues can be resolved not by gross appraisals
of a research structure, but by discerning evaluation of its individual
components.
The presentation will be divided into two parts: an outline of the
architectural structure and function of a clinical research project; and a
discussion of the special ethical issues raised by different elements of
different projects.
1 . The Architecture of Medical Research
Of the diverse designs that can be created in the world of the arts,
the plans made by an architect are the only ones that are regularly con-
strained by the harsh demands of reality. Since the creative activities
of medical research are similarly constrained by the realistic demands of
both science and people, we have used the word architecture , rather than
design , to refer to the intellectual structure of projects in medical re-
search.
9-2
1.1. The Basic Outline of the Objective
The customary starting point in an act of medical research is the investv
gator's statement of a question to be answered. This is the research objec -
tive . The quality of the ensuing research often depends on the clarity with
which this objective is defined. It is often expressed in the form of a
question: "Will B occur if X is done to A?" Examples of such questions
are the following research objectives: Will headaches be relieved if treated
with aspirin? Will this new vaccine prevent the development of poliomyelitis
in healthy children? In non-allergic people, will antibodies be provoked
by the intradermal injection of serum from patients with hay fever?
These statements each contain three distinct elements: (1) an agent or
process that is the principal maneuver under investigation; (2) a direct
or implied description of the baseline condition, or initial state , of the
person receiving the principal maneuver; and (3) a direct or implied des-
cription of the target event to be observed after the maneuver is imposed.
In the cited clinical circumstances, the principal maneuvers were, respec-
tively, treatment with aspirin, treatment with a new vaccine, and intra-
dermal injection of serum from patients with hay fever. The respective
initial states were headache, healthy child, and non-allergic person. The
respective target events were relief of headache, prevention of poliomye-
litis, and development of antibodies.
The main strategy in analyzing the ethical properties of a particular
research architecture is to consider the various maneuvers and procedures
that are imposed upon the people being studied. The principal maneuver
9-3
is the agent or process that the investigator is most interested in study-
ing. In the course of clinical research, however, as well as in ordinary
clinical practice, many different agents or procedures, i.e., maneuvers,
are also employed. Proper assessment of a medical research project re-
quires that attention be paid to the types of maneuvers imposed; the pur-
poses for which they are imposed; the background knowledge that sanctions
their acceptability, scope, and necessity; and the way in which certain
maneuvers are compared and assigned. These distinctive attributes create
both the architectural and ethical challenges of clinical practice or
clinical research.
1.2. Types of Maneuvers Imposed on Patients
All the procedures, processes, and agents that can be performed on
patients fall into one of two categories: action maneuvers or examinative
maneuvers.
1.2.1. Action Maneuvers
An action maneuver is intended to produce a persistent or temporary
change in a person's condition. When an action maneuver is imposed, the base-
line or initial state of its recipient is expected to change. There are two
types of action maneuvers.
1 .2.1 .1 . Interventional -'Maneuvers
An interventional maneuver is directly intended to produce a beneficial
change in its recipient. Aspirin administered to a patient with headache is
an interventional maneuver because it can give rise to a remedial improvement
9-4
in the patient's initial state, even though the improvement may only be
temporary. Vaccination of healthy children against poliomyelitis is
another example of an interventional maneuver. The intended change is
prophylactic, in contrast to the remedial change cited in the previous
example, but the initial state of the patient is modified. The maneuver
produces a more beneficial subsequent state, changing the patient, from
a "healthy" state of susceptability to poliovirus infection, to a pro-
tected state of immunity.
Diagnostic breast biopsy, on the other hand, is not an interventional
maneuver. Although knowledge of the results of a negative biopsy may pro-
duce benefit by relieving psychic stress, the maneuver itself (breast biopsy)
does not produce the change. For a maneuver to be interventional, it must
be capable of producing a beneficial effect by directly altering the state
of its recipient.
1.2.1.2. Explorational Maneuvers
An explorational maneuver is performed in order to produce a change,
usually transient, that is not itself expected to be beneficial. The
maneuver may give rise to information that ultimately benefits the recipient,
but the maneuver does not directly produce a beneficial change. A glucose
tolerance test, for example, is an explorational maneuver. The overnight
fast and initial glucose load create a transient physiologic change in the
state of the patient. Although the change itself is not beneficial, the
patient may later be helped considerably by the information that the glu-
cose tolerance test provides.
9-5
The intradermal injection of serum from patients with hay fever into
non-allergic individuals is another example of an explorational maneuver.
A change of state is expected, but the new state is not expected to be an
improvement. In contrast to the glucose tolerance test of the previous
example, the maneuver here is not expected to yield information that will
be promptly beneficial to its recipient.
1.2.2. Examinative Maneuvers
In contrast to interventional and explorational action maneuvers,
examinative maneuvers are not expected to produce a significant change in
the recipients. The maneuvers consist of things said or done to a per-
son in order to observe a response as in history taking or a part of the
body, as in physical examination; to obtain a sample of fluid, tissue,
or excreta, as in venipuncture and urine collection; or to produce a tech-
nologic substance, as in electrocardiography or chest roentgenography.
All these examples are things that are done to people, without signifi-
cantly changing them. A GI series, on the other hand, is an explora-
tional maneuver, rather than an examinative one, because the introduction
of barium produces a distinct, although temporary, change in the initial
state of the patient, converting the alimentary tract from being nonradio-
opaque to radio-opaque.
1.3. Delineative Procedures
In contrast to the action and examinative maneuvers just discussed,
which all involve the person to whom they are applied, delineative proce-
dures are not applied to people at all. These procedures consist of the
9-6
appraisals, analyses, and transformations performed on the specimens, des-
criptive data, and other entities obtained from people. Both action maneu-
vers and examinative maneuvers, however, are invariably accompanied by
delineative procedures. For example, the value of a blood hematocrit is
obtained by means of a delineative procedure performed on blood obtained
from a person by means of venipuncture, an examinative maneuver. Similarly
the determination of blood glucose levels in a glucose tolerance test, the
microscopic analysis of a surgically removed lung, and the intellectual
appraisal of a history and physical examination to arrive at a diagnostic
impression, are all delineative procedures based on the prior application,
respectively, of explorational , interventional, and examinative maneuvers.
Since the performance of a delineative procedure entails no risk or
discomfort to the patient, such procedures themselves create few ethical
problems with respect to either the subjects of human research or the
patients in ordinary clinical practice, who provide the raw material ex-
posed to the delineative procedures. The ethical aspects of delineative
procedures arise from the way in which the results of those procedures are
used. Results that can be misused or misinterpreted in a way that might
harm a patient create ethical concerns, but the procedures themselves are
ethically "neutral ."
1.4. Purposes of Maneuvers
The various types of action and examinative maneuvers described in sec-
tion 1.2. can be applied for three basic purposes: explicatory, clinical,
and preclinical .
9-7
1.4.1. Explicatory Maneuvers
A maneuver used for an explicatory purpose is intended to illuminate
or explain a biological phenomenon. Such maneuvers are employed in physio-
logic, biochemical, pharmacodynamic, or other studies in order to help
explain some feature of the patient's condition, or a mechanism of the
maneuver itself, or both. For example, when a patient with diabetes melli-
tus receives an infusion of sulfate to study its effect on the renal excre-
tion of potassium, the sulfate infusion is an explorational maneuver employed
for an explicatory purpose. The immediate goal of the maneuver is not to
improve health, not to ameliorate a manifestation of diabetes, and not to
prevent a future diabetic vascular complication, but rather to help explain
mechanisms of diabetic renal function. Furthermore, the information derived
from the imposition of the maneuver is not expected to yield any immediate
benefit for the patient.
1.4.2. Clinical Maneuvers
Maneuvers performed for a clinical purpose are intended to provide
direct or indirect benefits for their recipients.
1.4.2.1. Diagnostic Maneuvers
The aim of a maneuver performed for a diagnostic purpose is to produce
information that will be useful in the clinical care of the person who re-
ceives the maneuver. The diagnostic process may require the imposition of
different types of maneuvers, such as the examinative maneuver of a blood
culture, the explorational maneuver of a BSP test, or the interventional
9-8
maneuver of antibiotic therapy in a patient with a fever of unknown origin.
1.4.2.2. Therapeutic Maneuvers
Maneuvers employed for therapeutic purposes are interventional. They
are performed to produce a beneficial change in their recipients. The goal
may be prophylactic, to prevent some adverse future event; or remedial, to
alter or remove some existing undesirable manifestation. The administration
of a vaccine anticipated to prevent smallpox is an example of a prophylactic
interventional maneuver, as is the procedure of retinal photocoagulation to
prevent blindness in appropriate diabetic patients. Appendectomy for
appendicitis and aspirin for headache are examples of remedial interventions,
1.4.3. Preclinical Maneuvers
Maneuvers undertaken for preclinical purposes are explorational or
examinative. Although not expected to benefit their recipients, they are
expected to lead directly to new interventional or explorational maneuvers
having clinical value. Like explicatory maneuvers, preclinical maneuvers
are intended to yield information useful to the investigator but not imme-
diately useful or helpful to the recipient of the maneuver. Unlike other
explicatory maneuvers, however, preclinical maneuvers are expected to serve
as a direct prelude to the development of diagnostic or therapeutic proce-
dures.
For example, in Phase I studies of new pharmaceutical agents, healthy
volunteers or patients receive the drug for the purpose of determining phar-
macokinetics, dose-response curves, acute toxicity, bioavailability, or pre-
9-9
liminary evidence of therapeutic safety. The purpose is preclinical because
of the hope that the successful completion of the Phase I studies will yield
the information needed to allow the new agent to advance to therapeutic ap-
plications. Similarly, the preliminary evaluation of a proposed new sero-
logic test for gonorrhea will not directly benefit those patients who under-
go the examinative maneuvers (venipuncture) that the study entails, but the
purpose of those maneuvers is preclinical because of the expected demonstra-
tion of diagnostic usefulness for the new test.
1.5. Multi-purpose Maneuvers
Most maneuvers are employed for a distinct single purpose. The aspirin
is given to relieve a headache or blood is drawn to determine a diabetic
patient's blood glucose, the sole purpose of the maneuver is usually clini-
cal. Occasionally, however, the same maneuver is used for multiple pur-
poses. A maneuver employed for a primary clinical goal may also interest
an investigator for explicatory purposes. Thus, the aspirin that a patient
received clinically for a headache may be simultaneously viewed as an ex-
plicatory maneuver that, with suitable examinative procedures, can help
explain the action of aspirin on platelets. When a blood sample is drawn
as an examinative maneuver for the clinical purpose of determining glucose
in a diabetic patient, the additional determination of blood glucagon in
the sample can make the examinative maneuver also serve an explicatory
purpose.
For these two examples, the purpose of the main maneuvers is clinical
from the point of view of the patient and his personal physician, but expli-
9-10
catory from the point of view of an investigator. These different points
of view can enable a single maneuver to have multiple purposes. The ethi-
cal justification for the imposition of a maneuver should always be first
considered according to the maneuver's primary purpose, as viewed by the
patient or his personal physician. Thus, in the cited examples, the pri-
mary purpose of both maneuvers is clinical.
In medical research activities, an investigator may sometimes add an
explicatory purpose onto a maneuver whose primary aim is clinical. For
example, an investigator who wishes to study the behavior of macrophages
in human lung tissue might arrange to receive specimens of such tissue
from patients undergoing pneumonectomy as clinical treatment for lung can-
cer. This sort of "piggyback" research can often circumvent the ethical
difficulties of using a maneuver whose primary purpose is explicatory.
The investigator, in this example, avoids the ethical problems involved
in asking volunteers to undergo open lung biopsy.
1.6. The Sanction of Maneuvers
A maneuver can be regarded as sanctioned (or nonsanctioned) according
to its general acceptance, scope, and necessity.
1.6.1. General Acceptance
A maneuver that is part of the ordinary, generally accepted procedures
of the medical armamentarium can be called an established maneuver . Among
such maneuvers are pharmaceutical agents, such as aspirin and digitalis;
surgical procedures, such as appendectomy; explorational maneuvers, such as
9-11
coronary angiography; and examinative maneuvers, such as venipuncture and
chest roentgenography.
Maneuvers become established in various ways. Some agents, such as
birth control pills, become generally accepted as a result of scientific
studies demonstrating their efficacy and relative safety. Other agents,
such as digitalis and aspirin, became established as a result of tradi-
tional use in medical practice. A third mechanism of acceptance is acade-
mic enthusiasm, which can often lead to widespread usage of new procedures
and therapies that have received neither rigorous scientific study nor
traditional usage. The academic approbation is not always vindicated.
A notorious example of mistaken academic enthusiasm was the widespread use,
25 years ago, of high-dosage oxygen for premature babies. Although rapidly
accepted, this new treatment was later shown to have little or no therapeu-
tic benefit, while often causing blindness due to retrolental fibroplasia.
The enthusiastic adoption of such maneuvers can often be traced back to a
mistaken physiologic rationale, in which the therapy "makes sense" and
follows logically from accepted (but erroneous) biological theories of the
day.
A non-established maneuver is a procedure or agent that is not generally
accepted in ordinary clinical practice. The use of a hypothetical new drug
such as "curitol," proposed as a treatment for baldness, would be a non-
established maneuver because the actual merits and dangers of "curitol" are
unknown. The use of thalidomide, on the other hand is now non-established
because of its known toxicity. Thus, maneuvers may be non-established be-
9-12
cause their consequences are unknown in ordinary clinical practice or be-
cause they are known and not accepted.
1.6.2. Scope of an Action Maneuver
The scope of an action maneuver is standard when it is established
and its primary purpose is accepted in ordinary clinical practice. Aspirin
used for relief of headache and penicillin for eradication of streptococcal
infection are examples of maneuvers used in a standard way. In both these
examples, the initial condition and subsequent target for which the maneuver
was employed correspond to the circumstances in which the maneuver is employed
in ordinary clinical practice.
An action maneuver is used for a nonstandard scope when an established
maneuver is employed for a condition or aimed at a target event that is
different from what would occur in ordinary clinical practice. Aspirin
used to prevent strokes in patients with transient ischemic attacks, and
digitalis used to treat patients with obesity, are examples of established
maneuvers used in nonstandard ways.
1.6.3. Necessity of Examinative and Explorational Maneuvers
Most of the examinative and explorational maneuvers employed in ordi-
nary clinical practice are regarded as necessary for the care of the patient.
Examinative and explorational maneuvers are warranted if their use is justi-
fied by clinical necessity. For example, electrocardiographic monitoring
during acute myocardial infarction is a warranted examinative maneuver be-
cause of its accepted, necessary role in the proper clinical management of
such patients.
9-13
In the course of medical research, and occasionally in ordinary practice
(especially at academic centers), maneuvers are imposed to obtain data that
may be useful for research or teaching purposes, although not necessary for
the clinical care of the patient. Such maneuvers can be called non-essential .
As an example, suppose an investigator, suspecting that aspirin affects the
metabolism of alkaline phosphatase, arranges to measure the serum alkaline
phosphatase levels before and after the clinical administration of aspirin
to patients with ordinary headaches. The aspirin treatment would be an act
of ordinary clinical therapy, representing the standard use of an established
maneuver. The acquisition of data about alkaline phosphatase, however, re-
quires the performance of venipuncture, which would here be a non-essential
examinative maneuver, performed to help explicate the mechanism of action of
aspirin.
1.7. The Arrangement of Maneuvers
Medical research often requires maneuvers to be administered in a way
that allows their effects to be compared. An important part of the architec-
ture of medical research is devoted to the arrangements for such comparisons.
1.7.1. Comparative Maneuvers
Although the principal maneuver is the agent or process in which an in-
vestigator is most interested, the events that follow its imposition are not
always caused by that maneuver. Consequently, investigators must beware of the
erroneous conclusions produced by post hoc ergo propter hoc reasoning if the
principal maneuver is used alone. The main method of avoiding such errors
9-14
is to study a comparative maneuver, or "control." By observing the results
that occur when the comparative maneuver is applied to the same initial state
in a different person (or in the same person, when feasible), the investigator
can more judiciously evaluate the accomplishments of the principal maneuver.
The choice of the agent or procedure employed as a comparative maneuver
depends upon the research objective. If the objective deals with absolute
efficacy *-- what does the principal maneuver accomplish? -- the comparative
maneuver is usually an inert entity that mimics the principal maneuver but
that lacks its main active ingredient. The comparative maneuver may thus be
a placebo, an injection of saline, a lotion composed only of the vehicle in
which an active steroid was dissolved, a sham surgical operation, or no treat-
ment at all .
If the research objective deals with relative efficacy -- does the
principal maneuver work as well as some other maneuver? -- the comparative
maneuver consists of an agent or procedure whose action is to be contrasted.
In this way, a new treatment may be compared against an established old
treatment, or two old established treatments may be compared to see which
is better.
The comparative maneuver may sometimes be applied sequentially to the
same persons who receive the principal maneuver. Such situations are called
cross-over studies. They are common in explicatory circumstances, where the
transient effect of the maneuvers is often brief enough to allow the same
person to receive all of the various maneuvers under comparison. Cross-over
studies can also be conducted for therapeutic purposes, but are applicable
9-15
only for a limited number of maneuvers and clinical conditions. In a parallel
study , which is used occasionally in explicatory and commonly in therapeutic
activities, one group of persons receives the principal maneuver and another
group, the comparative maneuver. Regardless of whether a parallel or cross-
over arrangement is used, the planned therapeutic comparison of a principal
vs. another maneuver is called a clinical trial .
1.7.2. Planned and Unplanned Comparisons
In ordinary clinical practice, different therapeutic decisions are made
in an ad hoc manner. When these decisions have produced substantial experience
with different maneuvers, an investigator may collect the data and perform a
research survey to contrast the outcomes of the several maneuvers. For exam-
ple, many of the studies of the therapeutic usefulness of anticoagulants in
myocardial infarction or of diverse treatments for cancer were performed by
reviewing the records and comparing the outcomes of patients in whom different
ad hoc therapeutic decisions had been made. The results of the maneuvers were
compared, but a deliberate comparison had not been planned when the maneuvers
were imposed.
In a research experiment on the other hand, the maneuvers under compari-
son are assigned according to a deliberate research design. In most modern
clinical trials, the allocation of either the principal maneuver or the com-
parative maneuver to a particular patient (or the choice of the sequence of
maneuvers in a cross-over study) is accomplished by means of randomization .
For example, if propoxyphene (DarvonW) is being compared with aspirin for
relief of headache in a clinical trial, randomization would ensure that each
9^16
patient has an equal chance of receiving either agent, by eliminating the
judgment, whims, and prejudices of the investigator when the assignment is
made.
1.7.3. Method of Observation
Although randomized allocation of maneuvers has been a major advance to
improve scientific validity in clinical research methods, another major scien-
tific improvement has been the use of techniques that help eliminate bias when
the effects of the maneuvers are observed. In the single-blind technique,
patients are kept unaware of the particular maneuver they are receiving. In
the double-blind technique, the exact identity of the imposed maneuver is
unknown to both the patient who receives it and the investigator who observes
and assesses the subsequent effects. Occasionally, as in anticoagulant
therapy, the regulation of dosage requires a knowledge of the treatment and
of certain pharmacologic effects. In such circumstances, a double observer
technique may be employed. One physician, aware of treatment, regulates the
dosage. The other observer, kept "blind," notes the other effects. No matter
how the double-blind technique is arranged, adequate safeguards must be esta-
blished to allow the physician to identify the maneuver if an untoward reac-
tion should occur.
1.8. Research vs. Ordinary Clinical Practice
The foregoing discussion has shown the similarity of many activities that
occur both in research and in ordinary clinical practice. By considering the
cited architecture for types, purposes, sanction, and arrangement of maneuvers,
9-17
the distinctions between research and ordinary clinical practice come into
sharper focus.
1.8.1. Characteristics of Ordinary Clinical Practice
Although all types of maneuvers -- interventional, exploratory and
examinative -- can be used in ordinary clinical practice, the primary purpose
of the maneuvers is always clinical. Action maneuvers are generally sanc-
tioned as established and standard; examinational and exploratory maneuvers
are generally sanctioned as established and warranted. Furthermore, in ordi-
nary clinical practice no plans are developed for deliberate testing of com-
parative maneuvers.
1.8.2. Characteristics of Research
As in ordinary clinical practice, all types of maneuvers can be used in
medical research. In contrast, certain other features are found almost ex-
clusively in research activities.
1.8.2.1. Explication
Maneuvers employed primarily for explicatory purposes are not a feature
of ordinary clinical practice. When a maneuver's primary purpose is expli-
catory, the activity is research, not practice; and many explorational man-
euvers are regularly used with primary explicatory purposes that are often
quite far removed from the goals of ordinary clinical practice.
1.8.2.2. Preclinical Maneuvers
Preclinical maneuvers also occur exclusively in research activities,
9-18
Because such maneuvers are direct preludes to clinical maneuvers, the situa-
tions in which they are used are generally closer to ordinary clinical prac-
tice than those of explicatory maneuvers.
1.8.2.3. Nonstandard Action Maneuvers
Established action maneuvers, when used for a nonstandard scope, are
generally part of a research activity, and are usually tested with a compara-
tive maneuver. Occasionally, however, established maneuvers are used in
ordinary practice in ways which can not yet be called standard, but are not
part of a specific research design. An example of such a situation is a
physician's prescription of aspirin to prevent recurrent transient ischemic
attacks.
1.8.2.4. Non-essential Maneuvers
Although frequently present in research activities, non-essential exami-
native and explorational maneuvers are occasionally used in ordinary clinical
practice, particularly at academic centers, for pedagogical purposes. The
number of non-essential maneuvers can often be minimized in research by using
warranted maneuvers for multiple purposes, as discussed in section 1.5.
1.8.2.5. Acts of Comparison
In the circumstances just cited, an act of research could be identified
by the way a particular single maneuver was employed. Other clinical acti-
vities become identified as acts of research not via the function of a single
maneuver, but via a comparison of two or more maneuvers. A pre-planned com-
parison is often, in fact, the distinction that separates many acts of medical
9-19
research from those of ordinary clinical practice. The introduction of a
pre-planned comparison can often transform into research something that would
otherwise be regarded as ordinary medical care. For example, a clinical
trial of medical vs. surgical therapy for stable angina pectoris is a com-
parison of two therapeutic interventions that are both ordinarily considered
as established, customary treatments. If allocated to individual patients
by ad hoc judgments, the treatments are acts of ordinary care. If allocated
by a pre-planned randomization, the same treatments are part of research.
This type of medical research, comparing one standard therapeutic maneuver
with another, often quite closely resembles ordinary clinical practice and
presents a minimum of inherent ethical difficulty. Comparisons that involve
explicatory or preclinical maneuvers, or that test clinical maneuvers for
nonstandard or non-established scopes, are further removed from ordinary
clinical practice and present greater ethical challenges.
2. Ethical Issues in Research
Since the main ethical difficulties in medical research arise from what
is done to the patient, why it is done, and whether it receives the patient's
consent, the research activities performed on patients require careful analy-
sis. The taxonomy that has just been described for maneuvers and other aspects
of research architecture can be a powerful tool in that analysis by providing
a clear "dissection" of the structure and purpose of diverse medical activities.
In this section we shall use that new taxonomy in brief discussions of several
types of ethical issues.
9-20
2.1 . Issues in the Purpose of Maneuvers and Procedures
The primary purpose for which a maneuver is employed has important
ethical implications. For example, giving a toxic agent, such as nitrogen
mustard, to a healthy volunteer for explicatory purposes might be regarded
as patently unethical (even with informed consent), yet the same agent
could be readily justified if given for clinical purposes to a patient
with Hodgkin's disease for clinical purposes.
2.1.1. Clinical Purposes
Of the diverse maneuvers used in research, the ones that are administered
for clinical purposes have the strongest resemblance to the activities of
ordinary medical practice. Thus, therapeutic maneuvers that are established
and used in a standard way should create no greater ethical difficulty in
research than they do in ordinary practice. Nonstandard and non-established
maneuvers present progressively greater difficulties because their potential
risks and benefits are progressively less well known. Persons exposed to
nonstandard or non-established clinical therapeutic maneuvers, therefore,
should be especially well informed of the possible consequences of their
exposure.
Similarly, when a diagnostic maneuver that is used in research is also
warranted, its performance is ethically justifiable in the same way as in
ordinary clinical practice. Because most diagnostic maneuvers present little
potential risk or discomfort, the ethical difficulty is usually small, even
if the maneuver is non-essential. On the other hand, maneuvers that entail
9-21
significant risk, such as cardiac angiography, may present thorny problems
if they are non-essential.
When clinical circumstances warrant their use, non-established diagnos-
tic maneuvers, such as transvenous myocardial biopsy, can be regarded in the
same way as other non-established clinical maneuvers. If their use is not
expected to provide information of importance to the patient, non-established
diagnostic maneuvers are not really diagnostic at all, and should be regarded
as having an explicatory purpose.
2.1.2. Explicatory Purposes
When the primary purpose of a maneuver is explicatory, it offers no bene-
fit to its recipients. Special attention must therefore be paid to the harm-
ful potential of such a maneuver. If the maneuver is established, such as
using caffeine to stimulate gastric acid secretion, the risks may be known
and the ethical import of the explicatory activity is easy to assess. Dif-
ficulties arise for maneuvers that are non-established. When the potential
risks can only be guessed, the investigator should be expected to amass as
much preliminary information as possible, including the results of animal
studies, about toxicity. Recipients of explicatory maneuvers should always
be fully informed in advance about possible untoward reactions.
2.1.3. Preclinical Purposes
Preclinical maneuvers ethically resemble explicatory maneuvers in that
the recipients are not expected to benefit. Since preclinical maneuvers,
however, are more likely than explicatory maneuvers to lead to immediate
9-22
clinical rewards, preclinical maneuvers can be regarded with slightly more
benignity. Nevertheless, their possible toxicities should be fully investi-
gated and disclosed in advance to all research subjects.
2.1.4. Delineative Procedures
Delineative procedures, not being performed on patients, can themselves
neither harm nor benefit patients. The only ethical safeguards needed for
delineative procedures arise from the use of the results. If a result can
affect a patient's clinical management or some other significant aspect of
life (such as eligibility for life insurance), careful precautions must be
taken to ensure that the procedure is performed and interpreted correctly,
that the results are kept confidential, and that the patient's consent is
obtained for any disclosure. On the other hand, if the results of a delin-
eative procedure will not affect the patient in any way, the performance of
that procedure does not create an ethical issue for that patient. To obtain
the patient's consent for performing the procedure is unnecessary. For
example, a patient's consent would not be required for a research patholo-
gist to examine macrophages taken from a lung that was removed by a clini-
cally justified pneumonectomy.
2.2. Issues in the Necessity for Research
By inquiring about the purposes, sanctions, and comparison of maneuvers,
we can determine whether they are being used for acts of research. The next
step is to inquire about the reason or necessity for the research itself.
9-23
2.2.1. The Need for Explicatory Maneuvers
Today's achievements in medical science largely depend on yesterday's
advances in the understanding of human biology. Our current ability to
transfuse blood, transplant kidneys, prevent poliomyelitis, alter cardiac
arrhythmias, and replace bones is the ultimate clinical consequence of pro-
gress for which the groundwork was laid by research using explicatory man-
euvers. Since no animal model can totally duplicate human biological phe-
nomena, explicatory maneuvers must often be tested in people to increase
the basic fund of medical knowledge that may eventually benefit all mankind.
With imaginative perception, however, investigators can sometimes take
advantage, for explicatory purposes, of maneuvers administered for other
reasons. Such research can be performed by grasping the opportunity to ob-
serve the results of maneuvers imposed by nature, by doctors in the course of
ordinary clinical care, or by patients themselves. For example, knowledge
of the pharmacokinetics of certain drugs has been greatly augmented by the
study of patients who receive these drugs in ordinary therapy or who deli-
berately take overdoses. The atomic bombing of Hiroshima and Nagasaki al-
lowed investigators to study the effects of intensive radiation exposure
in humans. The natural occurrence of certain inherited enzyme deficiencies
in people has led to important clarifications in understanding human bio-
chemical pathways. In all of these examples, an investigator who might have
been ethically unwilling or unable to ask volunteers to receive the studied
maneuvers could avoid the otherwise thorny ethical problems of such research.
9-24
2.2.2. The Need for Therapeutic Maneuvers
Therapeutic maneuvers are evaluated in medical research to establish
whether or not a treatment is safe and effective for a particular condi-
tion. Such research is necessary because therapeutic efficacy and safety
in people cannot be predicted with certainty despite elaborate scientific
theories and extensive animal investigations. Many "established" thera-
peutic maneuvers -- such as the bloodletting and purging of a century ago
and the excessive oxygenation of premature babies in the most recent past
— ultimately turn out to be ineffective or tragically unsafe. The test-
ing of new treatments and comparative testing of old treatments is needed
to protect all patients from the harm that might come if a relatively small
proportion of patients did not participate in research on therapeutic man-
euvers.
2.2.2.1, Transmutability of Efficacy
The efficacy of a therapeutic agent refers to the ability of that agent
to produce a particular result for a particular condition. In the research
that establishes the efficacy of an agent, its scope is determined by the
clinical conditions and the associated effects or target events for which it
is given. The research can establish the agent's efficacy only for the scope
that has been investigated. Suppose, for example, that relief of pain is
chosen as the only target event under examination in a clinical trial of
steroids vs_. placebo in treatment of peptic ulcer disease. Although the
steroids will probably seem quite effective for relief of pain, they will
not have been shown effective for healing the ulcer — a target that was not
9-25
tested. In fact, steroid treatment in patients with peptic ulcer disease
may regularly relieve pain while also leading to perforation, sepsis, and
death.
The issue of transmutable efficacy -- deciding whether an agent that is
demonstrably effectively for one scope can be safely regarded as effective
for another -- sometimes raises interesting problems in the evaluation of
ethics. When Beecher^ 1 ' in 1966 condemned as unethical an interventional
experiment in which penicillin was withheld in a group of patients with
streptococcal pharyngitis, he stated "it is known that rheumatic fever can
usually be prevented by adequate treatment of streptococcal respiratory
infections by the parenteral administration of penicillin." When the par-
ticular study Beecher cited was performed ten years earlier, parenteral
penicillin had been established as an agent that could eradicate strepto-
coccal pharyngitis, but a further scope had not been demonstrated. Doctors
did not know whether treatment of streptococcal infections with penicillin
would produce a different effect: prevention of rheumatic fever. Ironi-
cally, the reason that the antirheumatic prophylactic efficacy of penicillin
was so well known in 1966 was mainly because of the results found in the
wery study that Beecher attacked.
If therapeutic agents had an unbounded scope, so that one form of effi-
cacy could readily be transmuted into another without any uncertainties or
needs for additional scientific evidence, some of the prime therapeutic di-
lemmas of our era could be instantly eliminated. There would be no need to
ask pharmaceutical companies to show that an anorexigenic agent leads to
weight loss in patients with obesity. The demonstration of anorexic action
9-26
alone would suffice. There would be no need to compare the results of medi-
cal therapy to see whether a vascular graft, which obviously provides an
anatomic remedy for coronary artery obstruction, also reduces angina, pre-
vents cardiac deterioration, and prolongs life. Because efficacy, however,
cannot be transmuted, its attempted conversion will continue to create
major uncertainties. The clinical investigations needed to answer all ques-
tions would seem ethically justified by the uncertainties.
2.2.2.2. Investigations to Acquire "Soft Data "
The choice of the target events studied in the investigation of thera-
peutic maneuvers obviously has major importance. Unfortunately, the current
"scientific" obeisance to objective "hard data" has made most modern evalua-
tions of therapy depend on a choice of target events that consist of easily
measured reliable effects, such as death, size of X-ray shadows, and elec-
trocardiographic changes. The "soft data" that describe the uniquely human
clinical phenomena of sick people are usually deliberately ignored as being
too subjective to receive "scientific" attention. Applying this scientific
standard, investigators usually evaluate white blood count but not pain;
size of tumor but not ability to work; change in serum enzymes but not
change in family relations; quantity of survival but not quality of life.
Consequently, when a treatment is determined to be effective and safe,
little or no consideration may have been given to its effect on the personal
or symptomatic aspects of human illness. As a result, well-intentioned doc-
tors and patients may be misled into thinking that a therapy shown to be
scientifically effective is actually the best therapy. Failure to consider
9-27
"soft data" often produces the iotrogenic misery of toxic drugs, surgical
mutilations, and calamitous expense.
Because a therapeutic maneuver may be certified as safe and effective
after scientific research that ignored important "soft data," further re-
search may often be necessary to fill in the gaps about human suffering
or gratifications. A valuable ethical constituent of therapeutic effec-
tiveness may be the inclusion of important "soft data." Reliance on "hard
data" alone may often have all the advantages of scientific objectivity,
but almost none of the humanistic or scientific merits of a concern with
the total spectrum of a treatment's impact. Research that leads to the
widespread use and acceptance of a therapeutic maneuver might sometimes be
considered ethically remiss if crucial "soft data" were omitted from the
evaluation.
2.2.2.3. The Value of a "Negative Result "
When a clinical trial is successful, the results can be directly and
immediately used in clinical practice. If treatment A is shown to be
superior to treatment B, or vice versa, all patients can benefit as soon as
the new knowledge is disseminated. Furthermore, a patient who received the
"inferior" treatment during the trial might even be able to benefit, imme-
diately after the trial, by transfer to the treatment that his participa-
tion helped establish as superior.
Although an investigator beginning a clinical trial expects its results
to indicate a significant difference between the agents compared, even a
"negative result," i.e. , one that indicates no difference, can be valuable.
9-28
A clinical trial showing that a particular agent is therapeutically equiva-
lent to a placebo prevents that agent from becoming accepted, thus providing
a potentially valuable service to patients. For example, suppose several
preliminary reports reveal that an expensive new (but actually hazardous
and worthless) drug "curitol" cures the common cold. Without a clinical
trial to demonstrate the lack of efficacy, millions of people might other-
wise have been subjected to this costly and slightly risky agent.
A clinical trial showing that a new non-established agent is therapeu-
tically equivalent to an old established agent can be valuable in many ways.
The new agent may prove quite useful in patients who have had adverse reac-
tions to the established therapy. Differences in cost or convenience might
make the new agent preferable in certain patients. Furthermore, the new
agent might be found to be useful in patients for whom the established thera-
py is ineffective.
Thus, a clinical trial can be valuable even when the outcome is a "nega-
tive result." If the research is designed properly, patients who partici-
pate in it have a considerable chance of helping themselves and other similar
patients no matter what the outcome.
2.3. Communication with the Patient
A patient who participates in medical research has a unique relationship
with the investigator, depending in large part on the primary purpose of the
principal maneuver.
9-29
2.3.1. Clinical Maneuvers
A patient who receives a clinical maneuver in medical research is pri-
marily motivated by the desire for good medical care and by the chance of
receiving a possibly superior treatment. Such a patient regards therapeu-
tic promise as the main thrust of the research and perceives the imposition
of the principal maneuver (or comparative maneuver) as an act of treatment
given by someone who assumes the responsibility of being the doctor. This
type of patient wants the investigator to be more a good doctor than a
good scientist. Although aware of the experimental nature of the research
activity, the patient usually expects (and deserves) a traditional doctor-
patient therapeutic relationship with the physician who renders the clinical
care.
In ordinary clinical practice, the exchange of information between doc-
tor and patient is a crucial part of the therapeutic relationship. The doc-
tor's empathy, authority, and power of suggestion not only influence the way
the patient feels, but often influence the actual course of disease. The
doctor must therefore do more than passively convey information. He must
carefully choose what he says and how he says it. Should he tell a patient
immediately that he suspects cancer, or wait until the definitive diagnosis
is made? Most doctors would wait, and most patients would want them to wait.
When therapy is proposed in ordinary practice, most doctors do not inform
patients of all possible risks and side effects. The therapeutic relation-
ship presupposes that to a certain extent the physician will exercise his
judgment in the best interests of the patient. To do otherwise would vio-
late the spirit and value of the doctor-patient relationship.
9-30
An investigator who imposes a maneuver for clinical purposes may some-
times be placed in a difficult position. For the therapeutic benefit of
certain patients, the investigator, as a doctor, may want to withhold in-
formation that could have adverse effects on the patient's morale or psy-
chic status. Withholding information, however, leaves the investigator,
as a scientist, open to the charge that unappealing details were deliberately
withheld in order to ensure the patient's participation in the research.
Irving Ladimer'' ' has proposed that "explanation may be limited if
it can be demonstrated that risk or harm to the patient by such explanation
is proportionally greater than the risk involved in the proposed activity."
In contrary opinions, the belief is that willful withholding of information
in medical research -- as well as in ordinary practice — is never ethi-
cally justifiable.
It seems reasonable to suggest that patients who are to receive non-
established maneuvers (as described earlier) be fully informed about the
potential risks, discomforts, and inconveniences. The recipient of a non-
established maneuver is to some extent in uncharted waters. He deserves
to know the risks of the voyage before he agrees to embark on it. If the
investigator believes that it is not in the best interests of the patient
to know these risks, then it is probably not in the best interests of the
patient to participate in the study. Although the investigator may believe
that the risks are small, he cannot know that the risks are small if the
maneuver in question is non-established. Consequently, a patient asked to
take a potentially serious risk should be made aware of the existence of
that risk.
9-31
On the other hand, in certain circumstances an investigator may be justi
fied in withholding information from a patient about to receive an esta-
blished maneuver. For such maneuvers, the risks to which a patient will be
exposed have usually been ascertained. The investigator can have some con-
fidence in his belief that a particular risk is small. Furthermore, for
either standard or warranted maneuvers, the risks incurred are analogous to
the risks that would be incurred in ordinary practice. As in ordinary clini-
cal practice, the patient in a medical research project, who receives a man-
euver for clinical purposes, must rely to a large extent on the good in-
tentions and clinical competence of his doctor. Although communication
between doctor and patient should be as full and complete as possible, the
investigator will sometimes need to exercise restraint and judgment in the
best interests of the patient. As Campbell' 2 ' has pointed out: "The only
real protection for the individual lies in the scrupulousness, conscience,
and personal integrity of the investigator."
2.3.2. Explicatory Maneuvers
In imposing maneuvers for explicatory purposes, an investigator is not
engaging in therapy and may not need to enter into a therapeutic relationship
with the patients who are his subjects. An explicatory experiment can often
be conducted while the patient is under the basic care of some other doctor.
It is important, however, that the patient not confuse what is done for his
benefit with what is done for the sake of research. Although such confusion
is less likely to occur if the patient's basic medical care is supervised by
someone else, the investigator sometimes functions both as researcher and
as personal doctor.
9-32
m
A patient may thus be asked by his personal doctor to receive an ex-
plicatory maneuver as part of a research activity in which that same doctor
is the investigator. This dual activity by the same physician may easily
islead the patient into believing that he will in some way benefit from
the research. For such situations, it is especially important that the
explicatory nature of the maneuver be carefully explained to its potential
recipients,
2.4. Issues in the Design of Research
All the ethical issues just cited arise from the viewpoint of a person
exposed to a research maneuver. A different set of ethical issues can be
approached from the viewpoint of the investigator, the medical profession,
and the future patients who will be affected by the results.
2.4.1. The Research Objective
Research that puts human subjects at risk should be worthwhile doing.
Expertly planned, rigorously executed, scientifically impeccable research
can sometimes be applied to an objective so nugatory or poorly conceived as
to be almost worthless. A person should not be asked to participate as a
subject in an experiment that may risk his well-being if the experiment
will have little value for science or society. A major responsibility of
the investigator and of the ethical review committee is to ensure that
people are not devalued by being asked to participate in trivial or bad
research.
9-33
Before a research proposal reaches the stage of review for the ethical
issues it raises, the protocol should first receive thorough, critical ap-
praisal of its scientific goals, plans, feasibility, and significance. To
allow such appraisal, the statement of the research objective must provide
enough information to enable a scientifically sophisticated reviewer to
evaluate the proposed investigation. A confusing or unclear description of
the research objective will raise doubt about the scientific merit of the
proposal, while also obscuring its ethical merits.
2.4.2. The Validity of the Overall Design
The details of a research proposal translate the research objective
into a concrete plan of action. If the plan of action is poorly designed,
the objective will be met inconclusively or not at all. Although questions
of validity in research design are traditionally regarded as purely scien-
tific problems, scientific validity becomes an ethical issue when people are
asked to participate in an experiment. Poorly designed human research is
unfit for human subjects.
A research design that leads to specious conclusions may also be unethi-
cal because of the possible consequences of those conclusions. For example,
when an inadequately performed retrospective case-control study produces
false claims of toxicity for a well established treatment or for an act of
daily life (such as coffee drinking or breast feeding), the needless fear,
distress and other adverse psychic effects of the false report must receive
appropriate ethical consideration. Similarly, if a clinical trial mistakenly
purports superiority for a new therapy, future patients may be subjected to
9-34
the possible harm of an ineffective new agent, while simultaneously being
deprived of whatever benefits the old therapy might produce.
Furthermore, specious conclusions from badly designed human experi-
ments can be extremely difficult to countermand. Despite persistent doubts
about investigative architecture and methods, the research may not be re-
peatable because of its technical difficulties, its logistic problems, or
the ethical qualms that were created by the original, possibly erroneous
report. When the results of a large-scale human experiment are seriously
questioned, as in the controversy occasioned by the University Group Dia-
betes Program (UGDP) study/ '' ^' ' investigators and practitioners may
both be left in an ethical quandry. Should practitioners abandon a hereto-
fore established therapeutic agent because of results found in research
whose scientific design is unacceptable to many thoughtful clinicians and
epidemiologists? Should a new group of investigators attempt to repeat
the study, correcting its methodologic flaws, but exposing new groups of
patients to a therapeutic agent that may be harmful (if the first study was
fg)
correct)? As Kabat v ' has pointed out, "Is it ethical to do a study on
human subjects with a design such that one may come up with the wrong answer
or no answer?"
Because of many currently unresolved problems in the design of clinical
research, the value and reliability of a particular investigation cannot
always be assessed or predicted in advance. A simple research design may
seem adequate to meet a modest objective, but final results do not always
conform to initial expectations. Nevertheless, the data often tempt an
investigator to infer unwarranted but tantalizing conclusions. Such conclu-
9-35
sions sometimes achieve a high degree of acceptability, despite the absence
of the necessary additional studies. The increasing acceptance of such
conclusions decreases the likelihood that a valid scientific study will
later be performed. Investigators who doubt the veracity of accepted "facts"
or "conventional wisdom" may not be motivated enough to overcome the techni-
cal and ethical obstacles of repeating a human experiment.
2.4.3. Inaugural Randomization vs. Historical Controls
The risk of premature acceptability is especially high for new thera-
peutic maneuvers, because clinicians seeking better methods of treatment are
often tempted by laudatory reports of new therapies whose value has not been
conclusively proved. By the time an experiment is conducted to provide con-
clusive proof, the treatment may be so thoroughly accepted that the compara-
tive testing of a "control" group is regarded as unethical. This problem
has led some commentators, particularly Chalmers/ ' to advocate that new
therapeutic agents be inaugurated in humans only in the setting of a ran-
domized clinical trial. Chalmers has stated that uncontrolled early "pilot
studies" -- in which all patients receive the new treatment, with none get-
ting a concurrent comparative maneuver -- are potentially unethical because
of the risk of premature acceptability. According to Shaw and Chalmers, 1, '
"Years may elapse before members of the medical profession slowly realize
that the therapy is not as good as claimed. Meanwhile, hundreds of thousands
of patients have been poorly treated. The opposite also surely occurs --
promising therapies are discarded because they were not given a fair trial."
9-36
The advocates of inaugural randomization would disallow the "pilot" re-
search that may evoke unwarranted conclusions. By compelling the initial
use of all new therapeutic agents to occur in the setting of a randomized
clinical trial, Chalmers believes that valid conclusions are more likely
to be obtained, thereby serving both science and ethics.
The opposing viewpoint, for which Freireidv 8 ' has been a leading pro-
ponent, is that ethical and technical difficulties would prevent investiga-
tors from assembling the relatively large numbers of patients needed to con-
duct randomized clinical trials for each of the large number of new thera-
peutic agents that are constantly being developed. The alternative pro-
posal is to perform small pilot studies initially and to compare the results
found for the new therapeutic agents against the data of "historical controls"
with patients treated in the past with an old therapy.
This pilot-study-vs. -historical-control approach would eliminate the
ethical dilemma of an investigator who sincerely believes in the superiority
of an untried new therapy and who therefore feels reluctant to deny it to
eligible patients. On the other hand, in the absence of better methods for
demonstrating that the "historical controls" and the pilot-treated patients
were prognostically similar before treatment began,'"' the comparison of the
non-randomized therapies may produce distorted results because of unrecog-
nized bias in the choice of patients subject to the compared treatments.
The controversy, which is still unresolved, merits attention here be-
cause it illustrates how problems of research architecture and scientific
validity become intertwined with issues in ethics.
9-37
Since inaugural randomization cannot be feasible because it cannot
be conducted ubiquitously, and since current taxonomic classifications are
inadequate to create valid comparisons for "historical controls," neither
proposal alone seems likely to solve the existing scientific and ethical
dilemmas. A possible solution is to create improved systems of clinical
taxonomy, thereby allowing historical controls to be used for many minor
studies, while reserving randomized trials for major therapeutic uncer-
tainties.
3. Conclusion
The ethics of medical research depend upon the architectural compo-
nents of the research. These components have been described in a system of
classification that can greatly aid the intellectual "dissection" of re-
search according to the types, purposes, sanction, and comparison of medi-
cal maneuvers and procedures. The distinctions noted in these different
entities can help clarify the focus of issues in both the scientific and
ethical evaluation of clinical investigation.
9-38
REFERENCES
1. Beecher, H.K. Ethics and Clinical Research. New Engl. J. Med. 274:
1354-1360 (June 16), 1966.
2. Campbell, A.G.M. Infants, children and informed consent. Brit. Med.
J. 3: 334-338 (August 3), 1974.
3. Chalmers, T.C. Randomization of the first patient. Med. Clin. No.
Amer. 59: 1035-1038 (July), 1975.
4. Feinstein, A.R. Clinical biostatistics: III-V. The architecture
of clinical research. Clin. Pharmacol. Ther. 11: 432-441, 595-610,
and 755-771, 1970.
5. Feinstein, A.R. Clinical biostatistics: VIII. An analytic appraisal
of the University Group Diabetes Program (UGDP) study. Clin. Pharmacol
Ther. 12: 167-191 (March-April), 1971.
6. Feinstein, A.R. Clinical biostatistics: XIV. The purposes of prog-
nostic stratification. Clin. Pharmacol. Ther. 13: 285-297 (March-
April), 1972.
7. Feinstein, A.R. Clinical biostatistics: XXVI. Medical ethics and
the architecture of clinical research. Clin. Pharmacol. Ther. 15:
316-334 (March), 1974.
8. Gehan, E.A. and Freireich, E.J. Non-randomized controls in cancer
clinical trials. New Engl. J. Med. 290: 198-203, 1974.
9. Kabat, E.A. Ethics and the wrong answer. Science 189: 505 (August
15), 1975.
10. Ladimer, I. Social responsibility in clinical investigation. Med.
Sci. 18: 32-41 (October), 1967.
11. Shaw, L.W. and Chalmers, T.C. Ethics in Cooperative Clinical Trials.
Presented at the Conference on "New Dimensions in Legal and Ethical
Concepts for Human Research," May 19-21, 1969, Sponsored by the New
York Academy of Science.
12. University Group Diabetes Program. A study of the effects of nypogly-
cemic agents on vascular complications in patients with adult-onset
diabetes. Part I: Design, methods, and baseline characteristics.
Part II: Mortality results. Diabetes 19: (Suppl . 2) 747-830, 1970.
9-39
10
HOW TO IDENTIFY ETHICAL PRINCIPLES
Alasdair Maclntyre, M.A.
Alasdair Maclntyre
An Essay
prepared for the National Commission for
the Protection of Human Subjects of Biomedical
and Behavioral Research on the subject of
How to Identify Ethical Principles: that is
1) How to recognize what arguments and
considerations, rules of actions, objections,
etc. can properly be acknowledged as "ethical",
and 2) How they are to be distinguished from
and related to, other kinds of arguments and
considerations (e.g. legal or technical,
esthetic or political) that bear on the
acceptability or inadmissability of human
actions.
10-1
What do we and what ought we to mean by 'ethical'? If we
go to the standard dictionary definitions we shall find in modern
English at least three distinct ways of using the word. First
of all and most importantly it is often simply a synonym for 'moral.'
Secondly it is sometimes restricted so that it is applied to moral
utterances only when they are viewed as expressions of principles
or bodies of moral theory. So 'ethics' is the name of systematic,
philosophical enquiry into the nature of morality. Thirdly,
'ethical' and 'unethical' are sometimes applied directly to conduct.
Usually although not always this third type of use has reference
to some special set of standards, more especially professional
standards.
In this essay I shall use the word 'moral' in place of 'ethical'
for the first type of use; and I shall reserve 'ethical' itself
for the third type of case. This is the way of using these words
that seems to accord most nearly both with ordinary English usage
and with the practice of recent writers on moral philosophy. The
strategy of my argument is as follows: first, I stress a seeming
paradox, namely that the use of evaluative expressions in general
and that of moral terms in particular seems to presuppose impersonal
standards in which rational men ought to be in agreement, but that
in fact there is wide disagreement both among ordinary moral agents
10-2
about standards and among moral philosophers about the definition
of morality. Secondly, I suggest that those two kinds of disagree-
ment are closely connected and that rival concepts of morality imply
rival contents for morality.
I then sketch two very different conceptions of what morality
is. Depending upon which of these we adopt - if either - we shall
mean two very different things by the words 'moral' and (in its
wider senses) 'ethical' . One consequence of this is that 'moral'
(or 'ethical') will point to quite different types of contrast
with 'legal' in each case. Another is that there turn out to be
two very different conceptions of the 'ethical' in the third and
narrowest sense.
My entire argument embodies two distinctive negative contentions.
The first is that it is impossible in our culture now to find a
systematic way of using such words as 'ethical' and 'moral' which does
not already embody not merely a particular morality, but a particular
contentious morality which is at war with its rivals. The second is
that because disagreements among moral philosophers parallel and
reflect the disagreements among moral agents themselves - moral
philosophers turn out to be merely the most articulate and systematic
examples of moral agents - philosphy cannot as of now resolve these
rivalries in any logically compelling way. The moral philosopher
cannot indeed, I believe, avoid partisanship, although he must hope
to be a fair and just partisan. Any notion of him as a neutral - analyst
who will merely dissolve confusion and unclarity, but not himself
qua philosopher take sides (a viewpoint dominant for many years in the
Anglosaxon world) is dangerously misleading. Part of fairness and
10-3
justice therefore is to warn you at the outset to be very cautious
in accepting what I say.
Finally in the third section of the essay I consider certain
aspects of morality which may be recovered independently of our stance
on the wider disagreements. What I emphasize is the lack of homo-
geneity in what is so recovered. That is, I stress the disparate
character of the different types of consideration which are perhaps
misleadingly grouped together under the rubrics 'moral' and 'ethical.'
10-4
What makes a moral reason for doing this rather than that a
moral reason? What makes a moral reason a moral reason ? The two
questions are inseparable and it is important to understand why.
Begin with a lexicographical point. In English and other natural
languages we can distinguish two quite different classes of
expression which may be used to answer the question "Why should
I do that?" The first type are sometimes called 'expressions of
personal preference'. They include "I want to do (or have) that"
"I prefer to do (or have) that" "I like..." "I approve..." and
so on. Note that they are all in the first person. The second is
the class of evaluative expressions - which includes distinctively
moral expressions 'good', 'right', the adjectives of the virtues,
such as 'just', 'generous' and 'courageous', and also 'beautiful',
'true', 'valid' and 'pleasurable'.
Consider a crucial difference in the import of these types of
expressions when they are used to answer practical questions. I
say to you, "Do such-and-such." You reply, "Why should I?" If my
response to your question is of the form "Because I want you to"
(or any other expression of personal preference) , then I give you
no reason for doing whatever it is independent of your relationship
10-5
to me. If you love me or hate me or fear me or are dependent on
my good-will in some way (for example, I am your superior officer),
then you have been given a reason for doing or not doing whatever
it is; but otherwise you have been given no reason. (If an
unidentifiable voice outside the room says "Do such-and-such"
and you ask "Why?" and the voice answers "Because I want you to,"
the only question that will produce a reason for doing whatever
it is, is "Who are you ?") But with evaluative expressions it is
quite otherwise.
If having said, "Do such-and-such," I respond to your question
"Why should I?" by saying, "Because it would be right" or "Because
it would be just," I purport thereby to give you a reason whose
force is completely independent of my being the person who happens
to have uttered it. The use of an evaluative expression is always
to point to some criterion or standard of value, the authority of
which is independent of the attitudes of both speaker and hearer.
The meaning of evaluative expressions is such that if they are
employed to give reasons in this way, then, if the reason is a
good reason for action, it would have been so even if the speaker
had never uttered it. Reasons which are expressions of personal
preference depend upon the context of utterance for their force
and authority; reasons which involve genuine evaluations do not.
It is important to stress that so far the point is purely a
linguistic one. It is part of the distinctive meaning of evaluative
expressions that they refer us to standards or criteria independent
10-6
of speaker and hearer; but what is referred to may not be there.
It is the distinctive meaning of the word "unicorn" that it names
a particular kind of animal; but there are no such animals. None-
theless, the point is not unimportant. For it rules out of court
immediately all those theories which have tried in one way or
another to analyze evaluative expressions as disguised expressions
of personal preference, attitude or feeling. It is still open to
us to conclude that no evaluative reasons have any force or
authority because there are no impersonal standards of value of
the kind required; but if we do treat any evaluative - and more
particularly - moral reasons as having force and authority, we
shall only be justified in so doing if we have been able to
vindicate the claims upon us of some impersonal source of authority.
Just this makes two features of our contemporary situation
striking. One is the inability of moral philosophers even to agree
on the meaning of the word "moral." In their useful anthology
G. Wallace and A.D.M. Walker list a number of possible defining
characteristics of morality discussed by contemporary moral
philosophers. Moral judgments, some have said, are necessarily
universalizable: if I maintain that morally I ought to do such-and-
such, then I am committed to maintaining that anyone else in
relevantly similar circumstances ought to do the same. Moral
judgments, some have said, are necessarily prescriptive; they
are injunctions to perform certain actions and anyone who said
"You ought to do such-and-such, but do not do it" - using "ought"
in its primary moral sense - would be contradicting himself.
Moral rules and principles, some have said, are just those rules
* The Definition of Morality , London, 1970.
10-7
and principles which are given overriding importance, or at the
very least are necessarily treated as very important. Moral rules
and principles, some have said, are those whose breach produces
certain specific types of response and sanction: remorse or
guilt in the agent , a particular kind of disapproval in others.
And finally some have argued that moral rules and principles are
those rules and principles which have a particular kind of content;
they are concerned with particular kinds of harm and benefit.
Every single one of these theses has been asserted and argued
for by more than one contemporary distinguished moral philosopher;
every single one of them has also been denied and argued against
by more than one such. I do not want to attend to the details
of their disagreements, so much as to the fact of disagreement.
For I take it that the inability of professional moral philosophers
to resolve disagreement about the concept of morality and the meaning
of such words as 'moral' and 'ethical' through argument is related to
the inability of ordinary moral agents to resolve their disagreements
about which moral principles are the correct ones. Consider two im-
portant contemporary moral debates.
I. A: A just war is one in which the good to be achieved out-
weighs the evils involved in waging the war and in which a clear
distinction can be made between combatants - whose lives are at
stake - and innocent non-combatants. But in modern war calculation
of future escalation is never reliable and no practically applicable
distinction between combatants and non-combatants can be made.
Therefore no modern war can be a just war and we all now ought to
be pacifists.
10-8
B: If you wish for peace, prepare for war. The only
way to achieve peace is to deter potential aggressors. There-
fore you must build up your armaments and make it clear that
going to war on any scale is not ruled out by your policies. A
necessary part of making this clear is being prepared both to
fight limited wars and to go not only to, but beyond the nuclear
brink on certain types of occasion. Otherwise you will not avoid
war and you will lose.
C: Wars between the Great Powers are purely destructive
and all of them ought to be opposed by revolutionaries ; but wars
waged to liberate oppressed groups and peoples, especially in
the Third World, are a necessary and therefore justified means
for destroying exploitation and domination.
II. A: Everybody has certain rights over their own person,
including their own body. It follows from the nature of these
rights that at the stage when the embryo is essentially part of
the mother's body, the mother has a right to make her own uncoerced
decision on whether she will have an abortion or not. Therefore
each pregnant woman ought to decide and ought to be allowed to decide
for herself what she will do in the light of her own moral views .
B: I cannot, if I will to be alive, consistently will that
my mother should have had an abortion when she was pregnant with me,
except if it had been certain that the embryo was dead or gravely
damaged. But if I cannot consistently will this in my own case,
how can I consistently deny to others the right to life that I
claim for myself? I would break the so-called Golden Rule unless
10-9
I denied that a mother has in general a right on abortion. I am not
of course thereby committed to the view that abortion ought to be
legally prohibited.
C: Murder is wrong, prohibited by natural and divine law.
Murder is the taking of innocent life. An embryo is an identifiable
individual, differing from a new-born infant only in being at an
earlier stage on the long road to adult capacities. If infanticide
is murder, as it is, then abortion is murder. So abortion is not
only morally wrong, but ought to be legally prohibited.
About these two arguments I want to make four major points.
The first concerns the systematically unsettlable and interminable
character of such arguments. Each of the protagonists reaches his
conclusion by a valid form of inference from his premises. But
there is no agreement as to which premises from which to start; and
there exists in our culture no recognized procedure for weighing
the merits of rival premises. Indeed it is difficult to see how
there could be such a procedure since the rival premises are - to
borrow a term from contemporary philosophy of science - incommensurable.
That is to say, they employ and involve concepts of such radically
different kinds that we have no way to weigh the claims of one
alternative set of premises over against another. In the first
debate an appeal to an Aristotelian concept of justice is matched
against an appeal to a Machiavellian concept of interest and both
are attacked from the standpoint of a Fichtean conception of liberation.
10-10
We have no scales, no set of standards, by which to assess the
v;eight to be given to justice thus conceived over against interest
thus conceived or liberation thus conceived. Similarly in' the
second debate an understanding of rights which owes something to
Locke and something to Jefferson is counterposed to a universalibility
argument whose debt is first to Kant and then to the gospels and
both to an appeal to the moral law as conceived bv Hooker, More,
and Aquinas .
Secondly, in this unsettlable character, in this use of in-
commensurable premises, these debates are clearly typical of moral
argument in our society. If the debates had been about euthanasia
instead of abortion or social justice instead of war, the characteristics
of the arguments would have been substantially the same. Perhaps not all
moral disagreement in our society is of this kind r but much is and the
more important the disagreement the more likely it is to have this
character.
Thirdly - and this is the point of my excursion into the
characteristics of moral disagreement - there are crucial links
between this kind of disagreement among ordinary moral agents over
which moral principles we are to adopt and the current disagreements
between moral philosophers about how morality is to be defined.
Indeed one not uncommon type of argument used by contemporary moral
philosophers has been of the form: if X's account of morality is
accepted, then such-and-such moral principles would be acceptable;
but those moral principles are precisely unacceptable, and therefore
X's account of morality must be rejected. (Examples are G.E.M.
10-11
Anscombe on some Oxford moral philosophers in'Vfodern Moral Philosophy,
( Philosophy , Jan. 1958), R.M. Hare on Max Black and myself in The
Promising Game" (in Theories of Ethics , ed. Philippa Foot, Oxford, 1967),
Beardmore on Philippa Foot (in Moral Reasoning , London, 1969). Appeals
to one's own interests of the kind that appears in B's argument in
the first debate are precisely not universalisable; appeals to
a law that exists independently of our apprehension and endorsement
of it, as in A's argument in the first debate and C's in the second
are precisely not prescriptive ; B's argument in the second debate
does not concern itself with specific kinds of harm and benefit in
the way that some have considered essential to morality; and all
these attempts to discriminate between some of these arguments as
belonging to morality and others as not would fail in the eyes of
those for whom morality simply consists of those principles which
their protagonists take to be overriding.
Thece is of course no simple and easy conceptual connection
between defining morality in one particular way and holding some
particular set of moral principles; the connections are complex
and sometimes indirect. Nonetheless a thesis has emerged: that
the conflict over how morality is to be defined is itself a moral
conflict. Different and rival definitions cannot be defended apart
from defending different and rival sets of moral principles.
This conceptual connection between the content of moral principles
and the definition of morality will perhaps be best elucidated by
considering its historical explanation. When I characterized the
rival moral premises of contemporary debate as Aristotelian, Machiavellian,
Fichtean and so on, I suggested something of the wide range of historical
10-12
sources on which contemporary moral argument draws , but I did so
by using the names of philosophers as a kind of allusive shorthand.
Three points need to be made in a more extended way. The first
is that the origins of contemporary moral debate are not to be
found only or even mainly in the writings of philosophers, but in
the forms of argument which informed whole cultures and which the
writings of philosophers articulate for us in exceptionally clear
and accessible ways: Aristotle is being treated here as a spokes-
man for at least a central strand in the culture of fourth century
Athens, Fichte as related in a similar way to nineteenth century
Prussia.
Secondly, the premises of contemporary moral debate have not
merely been inherited; they have also been torn from the social
and intellectual contexts in which they were originally at home,
from which they derived such force and validity as they possess. What
we have inherited are only fragments and one reason why we do not
know how to weigh one set of premises against another is that we do
not know what force or validity to grant to each of them in isolation.
Thirdly, as the conceptual connection between the content of
morality and its definition would lead us to expect, this fragmented
inheritance is embodied in our rival definitions of morality as well
as in our rival sets of moral principles. The extent to which this
is so will not be entirely clear from my initial summary of Wallace
and Walker's anthology; for they restricted themselves to the views
of one particular school of moral philosophers, those in the analytical
tradition which has dominated most Anglo-American philosophy in this
century. Yet there was some point in citing their anthology; for if
10-13
even philosophers who are all of the same philosophical cast of
mind to a large degree cannot agree on the definition of morality;
then much more widespread disagreement can be expected among
philosophies of different schools. And is to be found.
Fourthly, the central problem can now be set. I argued at
the outset that the peculiar function of evaluative expressions in
our discourse is to refer us to impersonal standards of value, to
give reasons whose force is independent of who utters them. The
implication is that in this part of our discourse we ought to be
able to arrive at rational agreements on central, if not always on
peripheral, issues. Yet the state of moral argument in our culture
shows this not to be so. We therefore seem to be in a dilemma:
either we have to reject the presuppositions of the dominant culture
of our own society or we have to reject the possibility of rationality
in moral argument. But the roots of this dilemma are, so I have
suggested, historical. It may therefore be worth asking for a
somewhat longer account of those roots in order to understand what
is involved in the dilemma.
II
The United States is the only known political society whose
self-avowed basis is assent not only to a particular set of highly
debatable moral judgments, but to an even more debatable philosophical
theory about the character of these moral judgments. Belief in self-
evident truths about equality and rights to life, liberty and the
10-14
pursuit of happiness require some very cogent philosophical basis.
Jefferson himself was clearly to some degree aware of this for he
cites as the sources of his thought - in the Declaration both the
"commonplaces" of his own age and the writings of four philosophical
theorists - Aristotle, Cicero, Locke and Algernon Sidney. What
Jefferson does not remark upon is the not merely heterogeneous, but
even mutually antagonistic character, of these philosophical sources.
The moral scheme represented by Aristotle and the moral scheme represented
by Locke may consistently both lack rational warrant, but they cannot
both possess it, for the truth of the one entails the falsity of
the other. The prospect looms: the United States is perhaps founded upon a
moral contradiction. If so, unclarity about what constitutes a
moral consideration may be central to our political culture. Perhaps
to be clear and coherent in this area is necessarily to be un-American.
Perhaps - but we need to explore further. To do so I am going to
ignore some of the stands in our moral culture which I have already
noted - the German Fichtean one, for example, and concentrate upon
the two suggested by Jefferson, although I shall not tie myself
specifically to the details of the thought of Aristotle or of Locke.
Instead I shall more generally delineate what I shall call the
classical view of value, a view shared in substance if not in detail
by many ancient and medieval thinkers, and the modern individualist
view. What we have inherited are parts of both and the whole of
neither and from this, I shall argue, derives the central incoherence
in our view of what constitutes a moral consideration.
10-15
On the classical view there is no concept of morality as we
distinctively understand it. There is indeed no word in classical
or medieval Greek or Latin which can he translated by our word
moral . It is true that our word moral derives etymologically from
the Latin moralis and that Cicero invented moralis to translate
the Greek n ciifo^ T the etymological ancestor of our word ethical .
It is also true that Liddell & Scott's Greek dictionary gives moral
and ethical as translations of h bl^QS . But it errs and etymology
turns out to be a bad guide to translation. ^ Ul:<o$ means having
to do with character and a man's character is the sum total of his
dispositions to act in determinate ways. Aristotle's Ethics is about
practical life in general. The excellences of character which he describes
go far beyond the limits of what most modern writers take to be morality.
Moreover for Aristotle ethics as a subject of enquiry is subordinate
to politics. But what he means by 'politics' is as different from what
we mean by 'politics' as what he means by 'ethics' differs from what we
mean. Politics is the enquiry into the proper form of the polis - and
we cannot translate polis by city-state if that suggests to us the kind
of contrast between the state on the one hand and the community on the
other which is central to political thought in our sense of 'politics.'
The polis is the community and the state, understood as one, not two.
The basic concept in the classical moral scheme is that of
the good for man . Men are conceived of as having an essential nature
informed by a '£/[oS . an end, movement towards which completes and
fulfils that nature, failure to achieve which frustrates and disappoints.
The human good is such that men cannot achieve it in isolation;
it is only as members of a community that men can achieve their good.
10-16
Hence the basic split between self-interest and altruism which
figures so prominently in modern moral thought is absent. I am
incomplete, not fully human, without others - my friends, my kin,
my city. Philoctetes, in Sophocles' play, describes himself as
having been rendered by enforced isolation on a desert island
"without friends, desolate, without a polis , a corpse among the
living," and this is in no way intended as a rhetorical exaggeration.
Without sharing it with others, there is no good to be achieved by
man; hence Aristotle's argument that one cannot be a good man
without being a good citizen.
What the good for man is is to be determined by rational argument.
There are of course many goods in human life and each of them is a
characteristic object of human desire: health, knowledge, honor,
prosperity, pleasure and so on. The good for man is that form of
life in which each of these goods have their due place; 'happiness'
is the name given to enjoyment that accompanies and is internal to
the activity of leading that form of life. 'Happiness' then is not,
on the classical view, the name of a separable psychological state.
It does not stand to the activities which it accompanies as end to
means, or as byproduct to achievement, or in any kind of purely
contingent relationship. The-happiness-of-climbing-a-mountain or
the-happiness-of-playing-tennis-well are not psychological states
superadded to the activity of climbing a mountain or of playing
tennis well; they simply are these activities as they are enjoyed
by one fully engaged in them.
Human excellences or virtues are those dispositions which
enable a man to achieve the good. A crucial excellence is that of
10-17
friendship. From the modern point of view it is odd to discover
friendship in the same list of the virtues that includes courage,
generosity and justice. But friendship is on the classical view a
necessary part of achieving the good; indeed friendship provides
the true form of relationship both between citizens and between kin.
Although friendship will characteristically involve what we shall
call affection and other psychological states, it is primarily a
matter of commitment and it is the precisely defined character of
the commitments which ensures that friendship itself is a precise
notion. Consider by contrast how loose and imprecise the notion
of friendship has become in our own culture.
The virtue of justice in the classical scheme presupposes the
general applicability of a concept of desert, of what is due to
each kind of person and to each person. Desert is in part a matter
of the place of each kind of person in the life of the polis and in
part of the past behavior and achievement of individual persons.
Justice is the first virtue of a community because without a just
sharing of goods and apportioning of punishments the common pursuit
of the good for man will not be served.
I have picked out the virtues of friendship and of justice for
special attention because we too possess conceptions of friendship
and of justice, but the dominant conceptions amongst us differ
markedly from the classical ones; and this is more likely to bring
out the nature of the difference between the classical scheme and
its modern successors than is attention either to those virtues
10-18
where our own view and the classical is much closer - as in the
case of courage - or to those classical virtues, such as great-'
souledness of which our culture scarcely has any conception. But
all the virtues in the classical scheme have the same function;
their cultivation is to result in putting human desires into an
order so that in seeking what we want we shall achieve and not
frustrate our human nature. Our desires and emotions pull us
in many directions initially and especially when we are young; moral
training is a matter of guiding them and redirecting them, so that
we avoid destructive excess in any particular aspect of life. But
it is wrong to read such classical maxims as "Nothing in excess" as
merely negations in their intention. There is no ultimate contrast
in the classical scheme between morality and desire, as there •'s
in some later schemes - the Kantian for example. Indeed it is
through moral training that we learn to fulfil desire, but in a
distinctively human way. Those who insist on setting no limit
to any one of the desires do not make themselves happy, but become
slaves to that desire - and they destroy the community of the polis .
The name for this vice is "7; AgovE.sK, (pleonexia) .
What one has to learn therefore is judgment: how to exercise
the relevant virtues in a particular situation. This capacity is
itself a virtue, the virtue of prudence. The capacity for judgment
in the classical sense goes strikingly undiscussed by modern moral
philosophers. They have tended to suppose that moral reasoning is
either a matter of skill in designing the means to achieve a given
end or a matter of bringing a particular case under a general rule.
10-19
■
Indeed commentators on Aristotle's account of practical reasoning
have characteristically misread it, by reading into Aristotle a
modern view of such reasoning and not seeing there what is
distinctive and unfamiliar. Modern moral philosophers, like modern
moralists, tend to lay a great emphasis upon rules . This emphasis
is absent in the classical scheme, for in that scheme it is held
that in crucial and central cases where moral judgment has to be
exercised the accepted maxims of morality by themselves will yield
no answer (just as for a judge the crucial precedent-setting cases
of judgment are those in which the accepted laws by themselves give
no answer) .
Finally it should now be clear that the presuppositions of the
whole classical scheme is b oth that men are able to and do achieve
rational agreement in what the good for man is and on how it is
to be achieved and that men are capable of governing themselves
both individually and collectively by reason. This view of man's
essential nature is perhaps what divides us most strikingly from our
classical ancestors. For it is not just in our moral transactions,
whether theoretical or practical, that we presuppose the untruth of
the classical view of human nature; it is also in the whole range of
the human sciences. Whatever the wide range of disagreements in
such disciplines as sociology, psychology and anthropology they
agree - and have agreed since the eighteenth century - in rejecting
any classical and more particularly any Aristotelian view of man.
This agreement does not by itself settle anything; for in order for
us to be rationally influenced by it we should have to ask whether it
results from the findings of those sciences or from assumptions that
10-20
their practitioners have made prior to these findings which result
in them interpreting what they learn in one way rather than another.
But whichever of these alternatives holds, the fact remains that the
classical view is massively inconsistent with a great deal that al-
most everybody now believes. At the same time we have inherited a
large portion of our moral vocabulary from the classical view.
For the classical view comes down to us not only from the Greeks,
but as mediated by various forms of theistic classicism - Christian
and Jewish Aristotelians and Platonists of a variety of kinds. Of
course I have therefore oversimplified in speaking of the classical
view; but at least the oversimplification has provided some conception
of a perspective which was once near universal in cultures that are among
the ancestors of our own.
The same complaint of oversimplification may with equal justice
be made against the sketch of modern individualism with which I now
want to complement my account of the classical view. Let me begin
from two starting points. 3he first is that the concept of rights -
which we take for granted so easily - only emerges in the modern age.
There is no expression correctly translated by our expression "a right"
in either classical or medieval Latin or Greek or Hebrew, and in this
we may note it is Latin and Greek and Hebrew and not modern English
or French which speak for the majority of the human race. There is no
expression in Japanese, for example, correctly translated by our
expression "a right" until nineteenth century liberal Japanese
political writers invented one. It is an interesting paradox that
those eighteenth century writers such as Jefferson or Robespierre
10-21
who believed that they intuited timeless truths about the rights
of man did so in a vocabulary that had historically come into
existence as a child of late medieval legal usage and which does
not seem to be found in the precise senses in which they used it
until a hundred and fifty years or so before their own time. But
it is easy to understand why it did emerge as a central moral as
well as legal concept. The central preoccupation of both ancient
and medieval communities was characteristically: how may men
together realize the true human good? The central preoccupation
of modern men is and has been characteristically: how may we
prevent men interfering with each other as each of us goes about
our own concerns? It is interesting that when John Stuart Mill
translates pleonexia he mistranslates it as "the desire to engross
more than one's share of advantages" - the best modern translation
is Nietzsche's "mehrundmehrwollenhaben" - for this is yet more
evidence of the way in which the individualist view throughout its
history assimilates other conceptual schemes to its own by mis-
interpretation and mistranslation and thereby presents itself as
though it were morality as such.
The classical view begins with the community of the polis and
with the individual viewed as having no moral identity apart from
the communities of kinship and citizenship; the modern view begins
with the concept of a collection of individuals and the problem of
how out of and by individuals social institutions can be constructed.
The isolated, autonomous individual is in classical Greek social
thought the pathological specimen - Thrasymachus in Plato's fiction
and Alkibiades the traitor in fact to both his own city and to Greece
10-22
were as pure an example as any; for modern morality we begin from
just such an individual as the fundamental unit of moral and social thought
and practice. We may notice that the word "individual," used as
I have been using it, is itself a linguistic innovation in the period
which marks the origins of modern morality. Until the seventeenth
century the word "individual" is almost exclusively a technical term
of logic, contrasted with the word "class," another technical logical
term which is about to acquire a social meaning. Its predecessors
are "man" and "person". "Man" in its Greek and Latin synonyms is both
biological and social; "person" emphasizes role-status. What matters
is the systematic character of the linguistic invention necessary to
make both the modern moral vocabulary and modern moral theory possible.
How can a society or a morality be constructed out of a set of
individuals? The answers are different in different philosophical
authors: Hobbes has one view, Locke a second, Diderot a third, Hume
a fourth, Adam Smith a fifth. But every one of these authors faces
in the context of his own distinctive theory one and the same problem.
For each of them sees the institutions of social and political life as
external to the basic projects and purposes of individuals who are
each moved by their own desires. The notion of a common good which
is the object of each man's true desire has disappeared. The extent
and the direction of individual self-interest is differently assessed
by different authors. Hobbes does not agree with Locke, Hume in the
Enquiry does not agree with himself in the Treatise and Diderot was
capable of radical disagreement with himself within the pages of one
and the same book. But however they characterize self-interest their
10-23
problem is: how can the authority of moral rules or that of social
and political institutions be the outcome of the coming together
of basically self-intersted individuals? The answer in these
authors is psychological. Human nature must have some other original
component which mitigates and modifies what would otherwise have
been the effects of self-interested desire. In Hobbes it is the
mutual fear of death at each other's hands, in Hume it is a larger
view of our own interests reinforced by sympathy, in Smith an original
capacity for sympathy is the mainspring of morality. But what matters
is not the contrast of the solutions (every one of which fails) , but
the form of the problem. For it is the form of the problem which
dictates the content of morality. Morality concerns that arrangement
of social rules which prevents my preying upon you and vice versa ;
morality therefore is a form of restraint upon our desires, and both
morality and social institutions are means to a further end. They
are instruments designed to enable us to achieve fulfilment of some
of our desires by sacrificing others. Hence the content of morality
becomes essentially a set of prohibitions framed as rules. The
further question at once arises; in what circumstances, if any, may
we make exceptions to the rules?
One striking fact about the moralists of the eighteenth century,
in France, England and Germany is that they are in fact highly con-
servative about the content of morality. With minor exceptions they
accepted the inherited set of rules, but understood them in a new
individualist way. But in accepting the inherited set of rules, they
accepted rules whose form was unamenable to understanding in the new
10-24
way for a reason on which I have so far touched only in passing.
Jewish, Christian and Islamic theology all inherited and nurtured
within themselves versions of the classical view of morality. But
they all added to it another set of elements. Man indeed has a
desire for his true end, even if that desire is corrupted by sin.
Man's natural rational morality is therefore congruent with divine
law since God created man to achieve his end by obeying that law.
But in his revelation in the scriptures God has set forth his law
directly; to the injunctions which spring from the virtues
the form and the authority of laws has now to be given. It is of
the essence of these laws that being divine they are unconditional,
that is, they admit of no exceptions and this character of un-
restricted universality and generality is written into the form
of the theistic moral rules. When these rules are inherited by
seventeenth and eighteenth century thinkers, they at once encounter
problems. We are to obey the rules of morality because they
restrain some of our desires and so enable us to fulfil others.
But what if on particular occasions continuing to obey the rules
will not serve the fundamental purpose of morality at all? What
if on this occasion breaking the rule will serve my interest? What
if on this occasion breaking the rule will serve the interests of a
number of people and harm nobody? Are there not occasions on which
we should break rather than keep promises, tell lies rather than the
truth?
We find therefore in the individualist account of morality a number
of elements which are not entirely coherent with each other. Morality
10-25
is to protect the invasion of my and your rights by others;
morality is to be the expression of my and your fundamental interests;
morality is to be a set of unconditional universal rules; morality
has no source but the will of each individual; morality is
essentially negative, a set of restraints on otherwise anarchic
and conflict-engendering desires. Confronted with these not entirely
compatible theses , which are increasingly embodied in the public and
private transactions of the seventeenth and eighteenth century, moral
philosophers responded by trying to impose consistency, often in the
name of some epistemological ideal or some view of human nature.
Kant, accepting both that the individual utters the moral law to
himself and that that law is expressed in universal and unconditional
rules, rejects the notion that moral rules are in any sense expressions
of our desires. Consistently he explains moral rules as an expression
of reason whose relationship to inclination is at best one of
neutrality, often one of constraint. Duty and inclination are to
be viewed as totally distinct. Smith by contrast constructs a
compromise between sympathy and self-interests; whereas the
utilitarians, rejecting any basis for morality except a psychological
one, find it difficult not to treat rules as rough and ready in-
structions on how to get as much as we can of what we want without
frustrating and being frustrated by other pursuers of desire.
We can already recognize in these mutually incompatible attempts
to articulate moral individualism into a consistent scheme part
of the ancestry of the rival contemporary accounts of what morality
consists in summarized by Wallace and Walker. Universalisability,
10-26
prescriptivism, and the view that morality is defined by its subject-
matter all have eighteenth century origins. Their detachment from
their origins has led their protagonists to make claims for them as
timeless, distinguishing marks of universal morality, rather than
as historical survivors from particular phases of moral history, a
phase itself marked both by incoherence and by a blindness to the
existence of its most important conceptual rival, the classical view
of morality.
The conclusion which I suggested earlier is then reinforced and
expanded by this abbreviated history: rival accounts of what morality
is presuppose rival -moralities and there is no way of deciding upon
the truth or falsity of the claims of any particular account of what
morality is or what we are to mean by 'moral' without deciding upon
the truth of falsity of the claims of the relevant morality. We may
well flinch from this conclusion because of the enormity of the task
which it seems to impose upon us. But in fact the situation is even
more difficult than I have suggested so far. For it turns out that
there are morally different ways of flinching, morally different ways
of responding to this situation.
Ill
The heirs of modern individualism have to recognize not only
their shared differences from the classical tradition, but also their
shared inability to agree with each other. They are apt to conclude
10-27
either that there is no such thing as the good for man or at the
very least that rational agreement on its character has to be
ruled out. Sir Isaiah Berlin writes that "the highest ends for
which men have rightly striven and sometimes died are strictly
incompatible with one another .. .Even if, per impossibile , we could
choose among these ideals, which should we accept? Since there is
no common standard in terms of which to grade them, there can be
no final solution to the problem of what men as such should aim
at" ( Vico and Herder pp. 211-212) and although Berlin is expounding
the thought of Herder, it seems clear that he is also endorsing it.
When John Rawls describes the way in which we ought to envisage
the terms of an original contract between individuals on which a
concept of justice might be founded, he says of those individuals
that they "do not share a conception of the good by reference to
which the fruition of their powers or even the satisfaction of
their desires can be evaluated. They do not have an agreed criterion
of perfection that can be used as a principle for choosing between
institutions. To acknowledge any such standard would be, in effect,
to accept a principle that might lead to a lesser religious or
other liberty..." ( A Theory of Justice , p. 327). Daniel Bell attacks
adherents of the classical view, of Catholicism and of Communism as
sharing a doctrine whose effect is "to fuse law and morality, to
insist that there is a single overriding principle..." and quotes
Berlin in support of moral pluralism. Berlin, Rawls and Bell are
interestingly different witnesses to the strength of the same position;
each sees the classical view of morality as not merely false, but
10-28
threatening. Each sees agreement on the good for man as something that
would necessarily be imposed rather than rationally shared.
We therefore have on their view to separate law from morality
and to separate that portion of morality which is necessary to under-
pin law if individual autonomy is to be safeguarded and its rights
acknowledged from any particular concept of the good for man. There
must be one part of morality which is private, various and reserved
for moral pluralism; there must be another which is public, and
where agreement can be secured without being imposed.
The classical view of morality does not allow for this division
of morality: law is to be at once the servant and the expression of
morality. The law is not merely an instrument to prevent one man
interfering with another to that other's harm; it is the community's
assertion to itself and to other communities of what it values and of
one means of achieving the good. It certainly does not follow from this
view either that everything held to be immoral ought to be made illegal
or that those who hold contrary views of morality ought to be legally
penalized. (Liberal thinkers such as Bell, Talman and Berlin have
often asserted a causal connection between totalitarianism and some
version of the classical view of man; but I would want to contend
that this is a highly dubious historical thesis. The roots of
totalitarianism lie elsewhere) . But it does mean that the contrast
law/morality appears quite differently in the individualist perspective
for the way it appears in the classical perspective; and this marks
one further variation in, one further fragmentation of the concept of
morality.
10-29
The classical perspective, with its demand for civic virtue,
experienced a certain rebirth in the middle and late eighteenth
century in the form of the revival of Greco-Roman and Italian re-
publican ideals. It coexists uneasily with individualism and
it is a clue to the ideological structure of the age that the conflict
between these two points of view largely went unrecognized. We have
already noticed this lack of recognition in Jefferson; it could
easily be parallelled in Robespierre and in many other lesser writers.
But this lack of acknowledgment is not restricted to the eighteenth
century. It is still at work unrecognized in contemporary argument where
what at first sight appear local and particular disagreements often
mark just this large rift between two rival total conceptions of morality.
Consider one particular case of just such a disagreement.
As I noted at the outset, the word 'ethical' has acquired a special
use in connection with professional standards. Within the history of
professions we can distinguish two quite different ways of envisaging
unethical conduct and they are in turn linked to two rival conceptions
of the character of a profession. One of these lays primary stress on
autonomy as the key characteristic of a profession. In virtue of their
shared exclusive possession of certain skills, the members of a pro-
fession are licensed by government to adjudicate on the legitimate use
of these skills. The public is entitled to ask that the profession
maintain standards; but it would infringe autonomy for the community
through government or any other agency to say what these standards
are to be.
By contrast there is an alternative conception which begins from
the classical notion of society as a community in which certain goods
10-30
are recognized as shared. Each good has its due place and because each
good cannot be pursued by everyone equally the community entrusts the pur-
suit of particular groups to particular groups of people: the goods of
health and life to the doctors, the goods of national independence to the
military, the goods of rational enquiry to schoolteachers and pro-
fessors. A profession is just a group to whom a special concern with
one particular type of good has been entrusted. Hence there is an
obligation laid on that group to cultivate and promote the skills
necessary for achieving the good in question. The autonomy accorded
to a profession by the wider community is a mark of the trust placed
by the community in that profession. It is in virtue of this agree-
ment upon goods and this trust that members of the professions are
accorded a type of authority and respect which goes far beyond anything
that could derive from the mere possession of skills. The authority
of physicians is perhaps the most obvious example; but lawyers,
professors, clergymen and others have in the past at least shared in it.
What has happened unfortunately is that this public agreement on
goods presupposed by this conception of a profession has been replaced
by moral pluralism and hence the concept of autonomy has become primary
instead of secondary. Hence professional standards lose their original
connection with more general beliefs about man and society and 'ethical'
takes on its special use. The claim to autonomy becomes an assertion
of a right rather than a sign of trust, a right of each profession to :
declare what is or what is not 'unethical.'
'Unethical' then always has reference to professional standards,
but it has been used to refer indirectly to more general standards in
virtue of which professional standards are to be vindicated. Once again
rival uses imply rival moral standpoints,
10-31
IV
Finally I return to the metaphor of fragmentation. The
range of disagreements which I have characterized has left us
not with a single unitary code or set of ideals, but with a
set of fragmented and often conflicting beliefs and concepts.
Often conflicting, however, but not always; and not all the
fragments we have inherited are of the same kind. We therefore
can set ourselves the task of distinguishing between what can
be rescued from our moral history and what cannot, what may
survive all disagreement and what certainly does not. I shall
deal in turn with four concepts: those of a virtue, a good,
a right - and more surprisingly perhaps - of a stranger. I shall
argue that we can rescue something important from the notions
of a virtue and a good, that the concept of rights is so gravely
damaged as to be unusable except in a legal sense and that the
concept of a stranger has for us something of the importance
which it had in Homeric times.
Begin then from the concept of a virtue. At first sight
this may seem a concept rendered almost unusable by the range of
disagreements concerning it. It is notorious that in different
times and places not only have different lists of the virtues been
given,but that different concepts of the virtues have been employed.
Aquinas following the Christian gospels treats humility as a virtue,
Aristotle treated it as a vice. Kant held that the very notion
of virtues (in the plural) was a confusion. And here once again
10-32
philosophers are representative voices of the wider culture.
Nonetheless, there is a core to the concept of the virtues and
there is a set of core virtues the acknowledgement of whose
authority over us is presupposed in any characteristically human
set of transactions. These are the virtues in terms of which we
define our relationship to each other whether we will it so or not.
Consider the case of truthfulness and imagine the situation of
three friends who have all been disturbed by and concerned over
the disappearance and obscure death of a fourth mutual friend.
A takes the trouble to discover precisely how and in what circum-
stances D died; A then tells the truth about D's death to B,
but lies to C. In so doing A has inescapably defined his relation-
ship to C as crucially different from his relationship to B. A may
intelligibly elucidate this difference further in a number of
different ways: he wished to protect a fragile and anxious C, but
considered B robust enough to endure the truth; or he wished to
protect D's reputation from gossipy C, but not from discreet B;
and so on. But A has defined C as having inferior standing to B
in his life in some respect and this he cannot intelligibly deny.
The norms of truthfulness provide the background in terms of which
we interpret the attitudes of others to ourselves and our own
attitudes to them. As with truthfulness, so also with fairness
and courage.
If in distributing rewards for merit or punishments for
offences I use any criteria other than those of merit and demerit
10-33
- for example, I give the highest grade not to the student who wrote
the best examination, but to the student with bluest eyes - I thereby
define my relationship to the students as other than that solely of
teacher to taught and it is by violating the canons of fairness
appropriate to the situation that I have done this. With courage
the matter is a little more complex. Courage is a virtue because
of the connection between caring for a person, a group, an institution
or an ideal and being prepared to risk harm or danger on their behalf.
The standard test for whether or how much I care or am concerned is
precisely what degree of discomfort, pain or trouble I am prepared
to endure to achieve the good of whoever or whatever it is. So we define
our loves, affections and loyalties by our will to endure; and without the
implicit acknowledgement of courage as a virtue they could not be so defined.
The virtues of truthfulness, fairness and courage are of course
embodied in very different codes in different cultures; but this
difference in codes is perfectly compatible with their functioning
with the same impersonal authority as defining notions in all those
cultures. Where we find a society, such as that of the Ik, in which
the authority of these virtues goes quite unrecognized we also find
a society lacking many other distinctively human characteristics;
it is in fact necessary to characterize the Ik in terms of
absence and loss, the absence and loss precisely of the core virtues.
The concept of a virtue functions in our evaluations very differ-
ently from that of a good. To elucidate the latter concept it is
important from the outset to distinguish what I shall call external
and internal goods. Some goods are internal to a given form of human
activity of practice. Example of such practices are mathematical
10-34
enquiry, agriculture, the common law and the game of chess. A
small child may originally learn to play chess because he or she
is being rewarded with candy both for playing at all and for
playing with increasing degrees of skill. So long as the child
is only playing chess to get the external good of the candy,
so long the child has no reason to excel at chess rather than merely
to give the appearance of excelling at chess. If, by cheating at
chess, the child can obtain the candy more or as easily than by
playing chess well, the child has indeed a good reason for cheating.
External goods are related to the activities which produce them
as end-products to means and economy of means is always rational.
But when the child comes to appreciate chess as an activity in and
for itself, when the child comes to desire those goods internal to
chess-playing, then the child by cheating merely frustrates him or
herself. For the child can only achieve the goods of chess-playing
by playing well. An internal good is not an end-product of its
activity; it is achieved in carrying the activity through to a
successful completion.
Many activities have both internal and external goods. Agriculture
as a mode of productive life in which man is related to nature in a
highly specific way has its external goods; but it also produces
turnips and money. Tennis is a game which like chess has its internal
goods; but it also produces health and sometimes money and reputation.
Yet external goods are always liable to corrupt participation in any
activity, while internal goods are not; and those who participate
together in an activity for the sake of the internal goods have a
10-35
quite different moral relationship from their relationship to those
who merely service or provision the activity in some way in return
for external rewards.
Judgments about virtues and judgments about goods are incorporated
by Aristotle and by many classical moralists into one overarching
scheme. But it is crucial to notice that, in the absence of any over-
arching conception of the good for man, judgments about virtues and
judgments about goods are logically independent of each other. This
is not just because they are judgments about different subject-matters,
virtues being qualities of individual agents' character and goods
being the end of activities and practices. For under certain circum-
stances - circumstances which arise often enough in our contemporary
social world - virtuous moral agents may destroy good and produce evil,
while men notably lacking in virtue may in fact achieve and create
notable goods.
It is just this type of situation which makes it practically
tempting to invoke the notion of rights, but also both theoretically
and practically dangerous tc a degree. For when the authority of the
core virtues and the authority of certain goods are as unrelated as
they are, we may well conclude that promoting excellence and avoiding
harm are not very closely connected. Virtues and goods are aims of
excellence; but if we are to avoid doing injury and damage we need
an additional and independent source of guidance - rules that will
constrain and prevent rather than enable and promote. It is in the
concept of a moral right that we are often tempted to search for a
foundation for such rules of negative constraint. But there is a
10-36
fatal objection to doing so.
This is the arbitrariness of all claims to rights. This
arbitrariness emerges clearly if we ask precisely what rights all
human beings possess, a question to which both the American and
the French revolutions professed to give an answer. To life, to
liberty, to happiness - to property, to a job, to a fair trial ...
what criteria qualify for inclusion in the list, what for exclusion?
The arbitrariness emerges once more if we ask under what conditions
each of our alleged rights is defensible. When right clashes with
right or with the public interest or with the greatest happiness of
the greatest number which is to give way to which? We only do not
drown in a sea of uncertainties by resorting to fraudulent assertion
and counter-assertion. Note that I am not suggesting that all claims
to moral rights at all times are arbitrary and unfounded - that was
Bentham's view. But claims to right have only not been arbitrary
when they have been asserted in specific historical contexts in
which a good was in danger of being sacrificed by an infringing power.
The claim to a right is then essentially negative in form. The
agents of the infringing power are told that they have no right to
deprive us of some good. The inference is from a premise of the form
"X has no right to deprive us of such-and-such a good" - a premise to be
established in turn by sceptical arguments - to a conclusion of the form
"Therefore we have a right to enjoy it." This was the form of the claim
in which the case against general warrants was argued both in English
and in Massachusetts courts in the eighteenth century. But this
intelligible, rational, context-bound claim was then generalized by
10-37
Jef fersonian philosophy so that the claim to natural rights is no
longer a conclusion derived from premises - but a somehow underived
claim.
It follows that considerations about rights are very different
in their rational force from considerations about either virtues
or goods; they may indeed have such force, but only in certain types
of context. If we insist on invoking rights outside such contexts
we shall misuse the notion in such a way that something like C.L.
Stevenson's emotive theory of moral judgments will become true;
claims to rights will function as mere expressions of personal pre-
ference, of attitude and emotion. The immediate result will be to
substitute assertion and counter-assertion for argument; the next
consequence most probably to discredit this form of discourse al-
together. (There are signs that this is already happening). Is there
then any other concept available to us in specifying the kind of
restraints on possible harm-producing acts which we need to accompany
any pursuit of goods and cultivation of virtues?
In the society reflected in the Homeric poems and in that reflected
in the Icelandic sagas - both earlier predecessor cultures of our own -
a crucial concept is that of the stranger. Both societies, unlike our
own, were made up of small communities in which everyone occupied a
clearly determined role, with both kinship and social relationships
specified as sets of mutual obligations. How to act and react in
relation to other members of the same community was in such a society
largely unproblematic, but the question obviously arose of how to react
to the stranger who arrives in or passes through the community and
who possesses no allocated role. About the stranger men do not know
10-38
whose kin he is or what his status may be and they cannot usually
verify what he tells them; nor do they know what goods he pursues
or what virtues or vices he possesses. They are not participants in
common activity with the stranger. All the cues that they normally
need to guide their responses are missing. They therefore have to have
a special code for this type of 'case.
Note that in such societies men do not treat strangers as
they do because of any phantom notion that strangers have rights.
It is because they themselves could not consistently pursue the
goods which they seek, or cultivate successfully the virtues to
which they aspire if they did not show a special kind of concern
for strangers, a concern embodied in rules enjoining hospitality and
safe-conduct, for example. When Odysseus comes upon the Cyclopes
the key test of whether they are civilized men or not is whether they
exhibit themis (a custom- and rule-governed attitude) towards strangers.
(The Cyclops, it turns out, eat strangers).
But what relevance can this concept, torn from its place in Homer
and the sagasj have in a society as different as ours is from theirs?
The answer is that in a crucial sense the relationship of most people
in our society has become that of strangers to strangers. It was not
so in the rural and small-town America of late eighteenth and nine-
teenth century America. But we outside our homes, our places of work
and our immediate neighborhoods are essentially strangers who have
learnt to meet each other with something of the same suspicions that
Odysseus had of the Cyclops (eating strangers can take a variety of
forms) . It is precisely a claim about the incompatibility of treating
10-39
others who are not participants in the same forms of activity as
objects for our consumption with our own pursuit of virtues and
goods which is embodied in the Homeric and Icelandic concept of
the stranger (the same word is used of both guests and strangers
in Greek) .
It is not just then that the surviving fragments of morality
are various and variously usable. It is also the case that if we
wish to use them it matters in what order we move from one to the
other. We can defend the virtues on one basis, the goods that we
pursue on another; but unless we already have provided a rational
account of those virtues and goods we prize, we shall have no
basis for invoking any such concept as that of the stranger. But
without such a concept we shall be at a loss how to act in many
types of situation.
What I am saying amounts to this: because what we have inherited
are a variety of fragments - sometimes mutually inconsistent and some of
them unusable by those committed to rational argument - we cannot
as philosophers content ourselves with the task of analysis. If we
did so, our reply to a request for an account of the concept of
morality, of the meaning of 'ethical', would have to be: there no
longer is such a concept or such a meaning. Instead we are forced
into a task of conceptual reconstruction. Any such reconstruction
will, as I suggested at the outset, itself be morally partisan.
Philosophers can no longer be comfortable in the claim that they
are only exhibiting in clear form the contents of the ordinary
agents' moral consciousness. If the view that I am urging is correct,
then it will be true that ordinary agents who believe that they already
10-40
possess a well-educated moral consciousness - as most people do -
are generally going to be people with a corrupted and deformed moral
consciousness. Hence this account, if true, ought to be unpopular.
A necessary condition of my arguments being acceptable by any is that
they should be rejected by most.
10-41
11
SOME ETHICAL ISSUES IN RESEARCH INVOLVING
HUMAN SUBJECTS
LeRoy Walters, B.D. , Ph.D.
SOME ETHICAL ISSUES IN RESEARCH
INVOLVING HUMAN SUBJECTS
LeRoy Walters
Center for Bioethics
Kennedy Institute
Georgetown University
"The Commission shall. . .conduct a comprehensive investigation
and study to identify the basic ethical principles which should
underlie the conduct of biomedical and behavioral research
involving human subjects. ..."
-Public Law 93-348
Summary: This essay is divided into an introduction, the main body of
the essay, and a conclusion. The introduction defines terms and delimits
the scope of the essay. In the main body of the essay seven major ethical
issues are discussed: (1) the moral justification for nontherapeutic
research involving human subjects; (2) research design; (3) risk-benefit
analysis; (4) the selection of subjects; (5) informed consent; (6) the
social control of research; and (7) the compensation of injured research
subjects. The conclusion attempts to synthesize the ethical principles
which have emerged from the preceding discussion.
Introduction
Human research (or human experimentation) can be defined as planned
manipulation, observation, or study of one or more human beings which differs
11-1
in any way from customary professional practice. Within the biomedical realm
three subcategories of research can be distinguished: (1) therapeutic
research, in which the design of the activity or procedure is solely to benefit
the patient — whether by prevention, diagnosis, or treatment; (2) nontherapeutic
research, in which the design of the activity or procedure is solely to gain
new knowledge; and (3) mixed research, in which the design of the activity or
prodedure is partially to benefit the patient-subject and partially to gain
2
new knowledge. This essay focuses on nontherapeutic research and mixed
research, that is, on research in which the therapeutic/nontherapeutic ratio
is less than 100/0. For the sake of brevity, the term nontherapeutic research
is used throughout the essay as a shorthand expression for the more cumbersome
phrase "nontherapeutic and mixed research"; both patient-subjects and subjects
are referred to simply as "subjects."
The primary emphasis in the essay is on biomedical rather than behavioral
research since a more clearly defined ethical tradition and a richer ethical
literature exist for the biomedical sphere. However, many if not all of the
ethical principles developed in the essay are also applicable to behavioral-
research activities.
The method of the essay is to analyze seven major ethical issues in
human research with a view to identifying one or more ethical principles which
are pertinent to each issue. Four of the issues — risk-benefit analysis, the
selection of subjects, informed consent, and social control — are derived
from the charge to the National Commission in Public Law 93-348. The issue of
research design is closely related to risk-benefit analysis and the selection of
subjects. The remaining two issues seem to the author to be of comparable
significance. They raise the questions: Should nontherapeutic human research
be performed at all? and, What debt, if any, does society owe to injured
research subjects?
11-2
At the beginning of each section relevant portions of three major codes
of research ethics will be quoted. These quotations from the Nuremberg Code,
the Declaration of Helsinki, and the 1974 DHEW Guidelines will serve to
indicate how much attention major traditional codes have devoted to the various
issues and will provide a point of departure for the further consideration of
each issue.
I. Moral Justification
"The experiment should be such as to yield fruitful results for the
good of society, unprocurable by other methods or means of study,
and not random and unnecessary in nature." Nuremberg Code, 1947, Rule 2.
"... It is essential that the results of laboratory experiments be
applied to human beings to further scientific knowledge and to help
suffering humanity. . .
"In the field of clinical research a fundamental distinction must be
recognized between clinical research in which the aim is essentially
therapeutic for a patient, and clinical research the aim of which is
purely scientific and without therapeutic value to the person subjected
to the research." Declaration of Helsinki, 1964, Preface.
"This review shall determine. . .whether the risks to the subject are so
outweighed by the sum of the benefit to the subject and the importance
of the knowledge to be gained as to warrant a decision to allow the
subject to accept those risks." Protection of Human Subjects, DHEW,
May 30, 1974 (italics added).
In the literature which discusses the ethics of human experimentation
surprisingly little attention is devoted to the moral justification of non-
therapeutic research. The modern codes which discuss human research, begin-
ning with the Nuremeberg Code, simply assume that some studies will be performed
"for the sake of the knowledge to be gained" or "for the good of society".
This lack of explicit attention to the problem of moral justification is
particularly striking when one considers that the traditional ethic of medicine
has been a patient-benefit ethic. The obligation primum non nocere has generally
11-3
been translated to mean "do nothing which is not intended for the direct
benefit of the patient". Because of this tradition patients have generally
4
come to expect what Charles Fried aptly calls "personal care". Viewed from
this perspective, nontherapeutic research is an innovation in the field of
medicine and represents a challenge to the traditional ethos of the physician-
patient relationship.
The major justification for human research advanced by the few authors
who have discussed the issue is a consequentialist argument — that the social
benefits accruing from human experimentation are very great or, conversely,
that the harms resulting from the cessation of such research would be extremely
grave. Robert Q. Marston, then Director of NIH, presented a cogent case for
this point of view in a 1972 lecture at the University of Virginia. He noted,
for example, the contrast between Dr. Benjamin Rush, who treated literally
hundreds of victims of yellow fever by purging and bleeding, and Dr. Walter
Reed, who in a carefully controlled nontherapeutic study demonstrated that
yellow fever is transmitted by mosquito bites. Turning to the present, Marston
observed that modern medical techniques are significantly more potent, for good
or ill, than earlier treatment methods and argued that the only alternative
to a plague of iatrogenic illness is the use of controlled clinical trials.
Thus, Marston seems to espouse not only the view that performing nontherapeutic
research confers benefits but also the opinion that not performing research
causes serious harm. This latter claim may be significant, since human beings
are generally thought to have only a weak duty to be beneficent but a much
7
stronger duty not to harm their fellow human beings.
However, even if the good-consequences argument is accepted as being both
accurate and important, it is possible that it should be overridden by other
11-4
considerations. For example, in the United States in cases involving
competent adults, the informed-consent principle functions independently as
a procedural check on the performance of all human research. If no adult
subjects are willing to consent to participate in a nontherapeutic research
project which promises major social benefits, the project will not be carried
out. By extension, it can be argued that if no adults in the society were
willing to consent to participation in nontherapeutic research, then no
nontherapeutic research involving adults would be performed.
If this analysis is correct, it reveals that in the United States
nontherapeutic research is considered to be a cesirable but not an essential
activity. Nontherapeutic research involving adults is able to proceed because
subjects can be found who are willing to take part in the research; however,
no adult persons are compelled to be research subjects. According to Hans Jonas,
this policy is ethically appropriate since progress, even medical progress, is
an optional goal and since the duty to participate in nontherapeutic research
Q
is not part of the social contract.
One can perhaps conceive of a national health emergency in which a
particular society would decide that nontherapeutic research were not merely
desirable but essential. In this extreme case, the informed-consent requirement
might be waived if no volunteers stepped forward, and the necessary number
of subjects would need to be drafted. Similarly, if no researchers were willing
to undertake the (possibly dangerous) required research, some researchers
would also be conscripted. For less apocalyptic times like the present, however,
a combination of consequential and non-consequential considerations seems more
appropriate. On the consequential side, the potential benefits of nontherapeutic
research are acknowledged to be important for the justification of such
11-5
research. On the non-consequential side, the promised benefits of nonthera-
peutic research are regarded as insufficient to justify the performance of
such research unless certain important human rights — for example, the
right of self-determination — are also respected.
II. Research Design
"The experiment should be so designed and based on the results of
animal experimentation and a knowledge of the natural history of
the disease or other problem under study that the anticipated
results will justify the performance of the experiment.
"The experiment should be conducted only by scientifically qualified
persons. The highest degree of skill and care should be required
through all stages of the experiment of those who conduct or
engage in the experiment." Nuremberg Code, Rules 3 and 8.
"Clinical research must conform to the moral and scientific
principles that justify medical research, and should be based on
laboratory and animal experiments or other scientifically
established facts.
"Clinical research should be conducted only by scientifically
qualified persons and under the supervision of a qualified
medical man." Declaration of Helsinki, Basic Principles 1 and 2.
If, as was argued in the previous section, nontherapeutic research can be
morally justified under certain circumstances, then it is possible to consider
how such justified research should be designed. The issue of research design
can be distinguished only with difficulty from the closely-related issues of
risk-benefit analysis and the selection of subjects. Excellent research
design helps to minimize risks to subjects, and every design for human
research obviously includes a method for selecting subjects. In this section
several general issues of research design will be discussed. The succeeding
two sections will treat some of the more specific problems associated with
risk-benefit analysis and the selection of subjects.
11-6
In recent years discussions of ethical issues in human experimentation
have tended to focus primary attention on the problem of informed consent. As
the quotations from the Nuremberg Code and the Declaration of Helsinki indicate,
however, the adequacy of research design has been an important emphasis in
codes concerning the ethics of human experimentation. In addition, several
recent essays have returned to the discussion of this traditional theme.
A negative principle which several commentators have emphasized is that
inadequately designed research should not be performed under any circumstances.
In the words of David Rutstein,
It may be accepted as a maxim that c poorly or improperly
designed study involving human subjects — one that could not
possibly yield scientific facts (that is, reproducible observa-
tions) relevant to the question under study — is by definition
unethical. Moreover, when a study is in itself scientifically
invalid, all other ethical considerations become irrelevant.
There is no point in obtaining "informed consent" to perform a
useless study. 10
According to this view, research which is inadequate scientifically is also
inappropriate ethically, since it exposes subjects to risk without a reasonable
hope of benefit to anyone.
More positively, adequate research design requires that, whenever possible,
laboratory and animal studies should precede the involvement of human subjects
in research. The National Commission strongly emphasized this point in its
recent deliberations, conclusions, and recommendations concerning fetal
research. A further principle, as the codes somewhat platitudinous ly note,
is that the investigators involved in any research project should possess the
requisite expertise and skills to carry out the research accurately and safely.
Three aspects of research design which have only begun to be debated in
the medical-ethics literature of the past fifteen years are randomization,
11-7
adaptive design, and the monitoring of double-blind studies. Between 1919 and
1930 R.A. Fisher developed most of the modern concepts of experimental design.
These concepts, which were derived from Fisher's agricultural field experiments
in England, stressed the importance of performing prospective randomized trials
in order to eliminate investigator bias and to achieve maximally-reliable
12
results. In part because of Fisher's influence, the prospective randomized
clinical trial has come to be regarded by many as the most authoritative method
of hypothesis-testing in biomedical research.
Because the allocation of subjects to medical treatments on a random basis
seems to contradict the personal-care ethos of medicine, a lively debate has been
carried on concerning alternatives to randomization and the appropriate circum-
stances for randomization. Alternatives suggested include the judicious use of
epidemiological data or the employment of various statistical adjustment methods
(e.g. , regression and analysis of covariance) to analyze the relative effective-
ness of treatments which have been assigned to various patient-subjects solely
on clinical grounds.
In discussing appropriate circumstances for randomization, even vigorous
advocates of prospective randomized clinical trials , like Thomas Chalmers , concede
that a prospective randomized trial should only be undertaken when the relative
effectiveness of alternative therapies is unknown.
1. If the clinician knows, or has good reason to believe, that a
new therapy (A) is better than another therapy (B) , he cannot
participate in a comparative trial of Therapy A vs. Therapy B.
Ethically the clinician is obligated to give Therapy A to each
new patient with a need for one of these therapies.
2. If the physician (or his peers) has genuine doubt as to which
therapy is better, he should give each patient an equal chance
to receive one or the other therapy. The physician must fully
recognize that the new therapy might be worse than the old.
Each new patient must have a fair chance of receiving the new
and hopefully, better therapy or the limited benefits of the
14
old therapy.^
11-8
For Charles Fried, even this careful formulation is inadequate, since it over-
looks the fact that a fully-informed patient might express a clear preference
for one of the alternative therapies.
Adaptive or sequential design has also been suggested by some commentators
as a device for minimizing risks to subjects. The general feature of such
designs is that they require periodic assessment of results in a prospective
trial and allow for the alteration or termination of the trial on the basis of
the information derived from already-completed parts of the trial. Specific
designs suggested by Weinstein are the two-stage study, pairwise sequential
■I £L
design, and the employment of an optimal decision rule. In the opinion of
Weinstein, "Whether or not randomization is chosen, adaptive methods should be
used as a matter of course."
Double-blind studies raise unique problems of research design. The plan
of such studies requires that the investigator forgo certain types of information
so that investigator-bias can be minimized as a factor influencing the results
of the studies. In forgoing the knowledge of which treatment is being admin-
istered to which subject, the investigator also loses the ability to monitor the
effects of the agents or procedures and to recognize when predetermined levels
of significance have been reached. These functions should be assumed by a
monitoring or safety committee which meets at regular intervals to assess the
18
progress of the study.
The general thrust of this section on research design has been to emphasize
a single ethical principle — the minimization of harm to human beings. Thus,
if the first section argued that nontherapeutic human research can be morally
justified in certain circumstances, this section has added the refinement
11-9
that careful, risk-minimizing research design is one necessary condition for
ethically-acceptable human experimentation.
III. Risk-Benefit Analysis
"No experiment should be conducted where there is a priori
reason to believe that death or disabling injury will occur; except,
perhaps, in those experiments where the experimental physicians
also serve as subjects.
"The degree of risk to be taken should never exceed that deter-
mined by the humanitarian importance of the problem to be solved by
the experiment.
"During the course of experiment the scientist in charge must
be prepared to terminate the experiment at any stage , if he has
probable cause to believe. . .that a continuation of the experiment
is likely to result in injury, disability, or death to the
experimental subject." Nuremberg Code, Rules 5, 6, and 10.
"Clinical research cannot legitimately be carried out unless
the importance of the objective is in proportion to the inherent
risk to the subject.
"Every clinical research project should be preceded by careful
assessment of inherent risks in comparison to foreseeable benefits
to the subject or to others." Declaration of Helsinki, Basic Principles
3 and 4.
"This review shall determine whether. . .subjects will be placed
at risk and, if risk is involved, whether (1) the risks to the
subject are so outweighed by the sum of the benefit to the subject
and the importance of the knowledge to be gained as to warrant a
decision to allow the subject to accept these risks....
"'Subject at risk' means any individual who may be exposed to
the possibility of injury, including physical, psychological, or
social injury, as a subject in any research, development, or related
activity which departs from the application of those established
and accepted methods necessary to meet his needs, or increase the
ordinary risks of daily life...." Protection of Human Subjects, DHEW,
May 30, 1974.
Any proposal for undertaking human research which fails to consider the
19
possible consequences of that research — including possible harms and benefits
is morally irresponsible. On the other hand, any proposal which does take
seriously the benefit-harm calculus is immediately faced with many of the quandaries
which have traditionally confronted utilitarian theories of ethics.
11-10
20
Several problems of risk-benefit analysis can be briefly sketched. An
initial difficulty is ascertaining the various agents who are likely to be
subject to risk, e.g., the subject, the subject's family, or society as a
whole. A further difficulty is the categorization of various types of risks
and benefits, e.g., physical, psychological, or social. When the categories
have been established, the identification of specific risks and benefits within
each category can proceed, e.g., increased life expectancy or mental depression.
The relative probability of occurrence of each potential risk can be projected
either in precise percentage-terms or in more general terms, e.g., 75% probability
or low probability. In addition, the likely duration of each anticipated risk
or benefit can be estimated more or less precisely, e.g. , for ten years or for
a brief period. A further complexity is the problem of commensurability ,
whether between agents, categories of risks and benefits, or specific types,
probabilities, and durations of risks and benefits. For example, can one
legitimately compare a potential risk to an individual subject with a potential
21
benefit to the society as a whole? Still to be addressed is the difficult
issue of collective choice , that is, if several persons undertake risk-benefit
analyses and arrive at differing conclusions, can one discover a mechanism for
22
aggregating the results of their individual assessments?
To the extent possible, risk-benefit assessments should be made objectively,
that is, available statistics on the mortality, morbidity, and other sequelae
of particular agents or procedures should be employed. However, it should be
recognized that many subjective factors are likely to be involved in any
assessment of risks and benefits, and particularly in any attempt to make
comparisons among agents or categories of risks and benefits.
Since no mere mortal has access to a master value-table on the basis of
which all risks and benefits can be authoritatively assessed and compared, the
11-11
most reasonable alternative would seem to be the establishment of a procedure
which insures that a variety of competing risk-benefit assessments will be
considered and balanced against each other. In particular, it would seem
essential to include at some point in the review process the viewpoints of
persons, both scientists and non-scientists, who are not institutionally committed
to the research enterprise as well as the opinions of persons who are specifically
charged with the responsibility of protecting the subjects' interests. From
this process of accommodation among divergent viewpoints should emerge a risk-
benefit assessment which, even if it is not objective, is at least less one-
sidedly subjective than the perspective of any individual participant in the
discussion.
What, then, is the role of risk-benefit assessment in determining the
appropriateness of human research? Risk-benefit analysis is closely akin to
23
judgments concerning "proportionality" in traditional ethical discussions.
It raises the question: Is a proportionate good likely to result if this
action is taken? Given the conclusions reached in Section I above concerning
the moral justification of nontherapeutic research, it would seem to follow
that a favorable (or perhaps a highly-favorable) benefit-risk ratio should
constitute a necessary condition for undertaking such research. However, as
will be argued in the subsequent section on informed consent, a favorable
benefit-risk ratio does not constitute a sufficient condition for proceeding
with research. Other conditions must be satisfied as well.
IV. Selection of Subjects
"...The person involved. . .should be so situated as to be able to
exercise free power of choice... and should have sufficient knowledge
and comprehension of the subject matter involved as to enable him to
11-12
make an understanding and enlightened decision." Nuremberg Code, Rule 1.
"The subject of clinical research should be in such a mental,
physical, and legal state as to be able to exercise fully his
power of choice." Declaration of Helsinki, Section III, Rule 3b.
" ' Informed consent ' means the knowing consent of an individual
or his legally authorized representative, so situated as to be able
to exercise free power of choice...." Protection of Human Subjects, DHEW,
May 30, 1974.
The codes of ethics which discuss human experimentation say very little
concerning the selection of subjects. Apart from the short general statements
cited above, the codes are silent.
The most extensive systematic discussion of subject selection appears in
Hans Jonas' well-known essay "Philosophical Reflections on Human Experimentation".
There Jonas proposes "identification" as the principle of recruitment in general.
This principle means that initially the recruitment of subjects should take place
within the scientific community itself, since the members of that community can
most fully identify with the goals and purposes of scientific research. Sub-
sequently, "...one should look for additional subjects where a maximum of
identification, understanding, and spontaneity can be expected — that is, among
the most highly motivated, the most highly educated, and the least 'captive'
24
members of the community." Negatively, Jonas' principle of identification
means that the scientific community ought to refrain from taking advantage of
the readiest sources of supply — "the suggestible, the ignorant, the dependent,
25
and the captive."
The question of experimentation involving ill subjects presents a special
problem for Jonas' position, since some patients may be eager to identify with
the goals of biomedical research but at the same time may be quite suggestible
and dependent. Jonas argues that the sick should be involved in nontherapeutic
11-13
research only as a last resort and then only if the experiment is directly
26
related to their own disease.
There is in the history of medicine an impressive tradition of auto-
27
experimentation which corresponds closely to Jonas principle of identification.
One thinks, for example, of Werner Forssman who performed the first cardiac
catheterization on himself, then calmly walked to the radiological department
28
for X-rays with the catheter still in position. On the other hand, there
is a growing body of evidence to indicate that a substantial fraction of the
nontherapeutic research currently being conducted in the United States involves
precisely those persons who, according to the principle of identification, are
the least eligible candidates for such research. Specifically, clinic or ward
patients and prisoners seem to be involved to a disproportionate extent as the
29
subjects of nontherapeutic research. In some nontherapeutic research projects
at urban medical centers, seventy to ninety per cent of the subjects are ward
patients. One sociological survey also indicates that qualitative discrimination
may accompany this quantitative overrepresentation. In a study of 352 research
projects, Bernard Barber and associates found that ward or clinic patients were
much more likely to be involved in research with an unfavorable risk-benefit
30
ratio than were their private-patient counterparts.
There are, of course, historical reasons for the location of many major
medical centers in areas of urban poverty. The design of some research projects
also requires stable or quarantined populations like those available in various
closed institutions. However, these excusing conditions should not deflect
attention from the fact that in contemporary American society the risks and
benefits of nontherapeutic research are inequitably distributed. Even if Jonas'
stringent standard of noblesse oblige cannot be reached, simple justice would
11-14
seem to require that each economic class and group bear a proportionate share
of research risks. Given the marginal nature of the sacrifices required in
most non therapeutic research, adequate numbers of subjects from each class and
group may be willing to volunteer. Affirmative action programs which seek to
involve more well-educated and well-to-do persons may also be successful. If
not, then the public will in effect have voted to assign a higher priority to
. . 31
other types of civic activity.
Equity in the selection of subjects (and its obverse, non-discrimination
against the sick, the prisoner, and the less fortunate) thus constitutes an
additional standard for nontherapeutic research. Like adequate research design
and a favorable benefit-risk ratio , it is a necessary but not a sufficient
condition for ethically-acceptable research.
V. Informed Consent
"The voluntary consent of the human subject is absolutely essential.
"This means that the person involved should have the legal capacity
to give consent; should be so situated as to be able to exercise free
power of choice, without the intervention of any element of force,
fraud, deceit, duress, over-reaching, or other ulterior form of
constraint or coercion; and should have sufficient knowledge and
comprehension of the subject matter involved as to enable him to make
an understanding and enlightened decision. This latter element requires
that before the acceptance of an affirmative decision by the experimental
subject there should be made known to him the nature, duration, and
purpose of the experiment; the methods and means by which it is to be
conducted; all inconveniences and hazards reasonably to be expected; and
the effects upon his health or person which may possibly come from his
participation in the experiment. . . .
"During the course of the experiment the human subject should be
at liberty to bring the experiment to an end if he has reached the
physical or mental state where continuation of the experiment seems to
him to be impossible." Nuremberg Code, Rules 1 and 9.
"The nature, the purpose, and the risk of clinical research must
be explained to the subject by the doctor.
11-15
"Clinical research on a human being cannot be undertaken without
his free consent, after he has been fully informed; if he is legally
incompetent, the consent of the legal guardian should be procured.
"The subject of clinical research should be in such a mental,
physical, and legal state as to be able to exercise fully his power
of choice.
"Consent should as a rule be obtained in writing....
"At any time during the course of clinical research the subject
or his guardian should be free to withdraw permission for research
to be continued." Declaration of Helsinki, Section III, Rules 2, 3a,
3b, 3c, and 4b.
"'Informed consent' means the knowing consent of an individual
or his legally authorized representative, so situated as to be able
to exercise free power of choice without undue inducement or any element
of force, fraud, deceit, duress, or other form of constraint or coercion.
The basic elements of information necessary to such consent include:
(1) A fair explanation of the procedures to be followed, and
their purposes, including identification of any procedures
which are experimental;
(2) a description of any attendant discomfort and risks reasonably
to be expected;
(3) a description of any benefits reasonably to be expected;
(4) a disclosure of any appropriate alternative procedures that
might be advantageous for the subject;
(5) an offer to answer any inquiries concerning the procedures; and
(6) an instruction that the person is free to withdraw his consent
and to discontinue participation in the project or activity
at any time without prejudice to the subject." Protection of
Human Subjects, DHEW, May 30, 1974.
Since Nuremberg no aspect of human experimentation has received greater
attention than the issue of informed consent. As the foregoing quotations
suggest, during the past thirty years the language employed to describe
informed consent has gradually been refined and made more precise.
Several commentators have noted that informed consent can be analyzed
into two component parts. The information-component refers to a reasonable
disclosure by the investigator and adequate comprehension by the prospective
subject. The consent aspect of informed consent refers to an uncoerced
decision by which the subject agrees to take part in the disclosed and
32
comprehended procedure or project.
11-16
A variety of justifications or rationales for the informed-consent require-
ment have been proffered. According to Paul Ramsey, "The principle of an
informed consent is the cardinal canon of loyalty joining men together in
medical practice and investigation." 33 Within the American legal tradition of
the twentieth century, major emphasis has been placed on protecting the right
of self-determination and the bodily integrity of patients (and, by implication,
of patient-subjects) . Court decisions reflecting this emphasis include
Schloendorff v. New York Hospital , decided in 1914, and three 1972 decisions,
34
Cobb v. Grant , Wilkinson v. Vesey , and Canterbury v. Spence . Other rationales
for the informed-consent principle are suggested by Alexander Capron : "to avoid
fraud and duress," "to encourage self-scrutiny by the physician-investigator,"
"to foster rational decision-making," and "to involve the public." J Thus, both
conseguentialist and non-consequentialist justifications for informed consent
have been advanced, with the latter predominating both in the relevant court
decisions and in the secondary literature.
Despite the impressive consensus concerning the significance and necessity
of the informed-consent principle, several specific topics continue to spark
vigorous debate. Three of these topics will be briefly noted: (1) additional
types of information which should be conveyed to subjects; (2) subjective vs.
objective standards for informed consent; and (3) exceptions to the informed-
consent requirements. Proponents of fuller disclosure to potential research
subjects have advocated that information concerning the following questions
should be provided to each subject prior to his or her enlistment in any
37
nontherapeutic research project: the use of randomization in the study-";
38
the existence of a placebo group ; measures which will be taken to insure
3Q
the early detection and minimization of harms to subjects ; and provisions
11-17
40
for the correction of, or compensation for, research-related injuries. In
addition, one commentator has proposed that during the course of a clinical
trial already-enlisted subjects should be informed of any findings of the
study which might affect their decisions to remain in, or to withdraw from,
41
that study.
The nature of the standards for informed consent seems to have been left
somewhat unclear by the three major informed-consent court decisions of 1972.
All three decisions agreed that a patient-oriented standard ought to be preferred
to the traditional "reasonable medical practice" standard. However, two of the
decisions seem to require that a physician (or physician-investigator) disclose
42
all material facts which a reasonable person would wish to know. The California
court, on the other hand, adopted a more subjective standard, arguing that
"The scope of the physician's communications to the patient. . .must be measured
by the patient's need, and that need is whatever information is material to the
43
decision." In the research context the choice of a reasonable-patient or
a subject' s-need standard will significantly affect the stringency of the
disclosure requirement.
Several commentators have also raised the question whether the informed-
consent requirement can be waived in any non therapeutic research situations.
The Kefauver-Harris amendments to the Federal Food, Drug and Cosmetic Act in
1962 seemed to allow for such an exception by requiring investigators to obtain
the consent of the subject or the subject's representative "except where they
deem it not feasible or, in their best professional judgment, contrary to the
best interests of such human beings." 44 However, the regulations implementing
this provision seem clearly to confine this exception to the situation in which
45
an unlicensed drug is administered to a patient solely for therapeutic reasons.
11-18
Robert Levine suggests that the purpose of some studies may need to remain
undisclosed if the goals of the study are to be accomplished; however, he
argues that in such cases the prospective subjects should be informed that
some information is being withheld from them until the conclusion of their
46
participation. Perhaps certain types of unseen-observer research are the
strongest candidates for exception from the informed-consent requirement. '
Even in these cases, however, unconsented nontherapeutic research — while
perhaps posing no direct risks to the unknowing subjects — may constitute
48
a violation of their right of privacy.
The issue of informed consent is thus a central topic in discussions
concerning the ethics of human experimentation. As noted previously in the
section on research design, informed consent alone is insufficient to justify
proceeding with research, particularly in cases where no useful purpose can
be served by an experiment. However, reasonably free and adequately informed
consent is clearly one of the necessary conditions for ethically-acceptable
human research.
IV. Social Control
"Safeguarding the rights and welfare of subjects at risk in
activities supported under grants and contracts from DHEW is primarily
the responsibility of the institution which receives or is accountable
to DHEW for the funds awarded .... In order to provide for the adequate
discharge of this organizational responsibility, it is the policy of
DHEW that no activity involving human subjects to be supported by DHEW
grants or contracts shall be undertaken unless a committee of the
organization Can Institutional Review Board! 9 has reviewed and approved
such activity, and the institution has submitted to DHEW a certification
of such review and approval, in accordance with the requirements of
this part." Protection of Human Subjects, DHEW, May 30, 1974.
The issue of social control differs significantly from the issues discussed
11-19
in Sections II-V above. From an ethical perspective, it seems to raise
procedural rather than substantive questions.
Within the professions generally and within the medical profession in
particular the preferred method of social control has been self -regulation.
In the context of clinical practice, where the physician's exclusive concern
was to benefit the patient, a system which relied heavily on professional
self-regulation may have proved sufficient to protect the best interests of
the professional's clients. In the nontherapeutic research context, however,
the welfare of the subject ceases, by definition, to be the (medical or non-
medical) professional's exclusive concern.
In the United States the primary mechanism for the social control of
publicly-supported human research has been peer review of proposed research
projects by expert scientists and clinicians. These expert reviewers have
fulfilled a significant role, particularly in the assessment of research design
and risk-benefit ratios. Within the highly-developed NIH review mechanism, the
peer-review findings of the study sections have been complemented by the
52
judgments of the Division of Research Grants and the Advisory Councils.
During the decade of the 1960's an additional mechanism for the control of
human research was developed. Early in the decade, and even during the late
1950' s, local peer review committees began to be established, at first in a
small minority of medical centers. Later, when the Public Health Service
began to require prior review as a condition for grant support, the number of
such local committees multiplied rapidly. By the end of 1969 local peer
review committees were, in the words of Bernard Barber, "probably universally
53
established in biomedical research institutions."
During the 1970 's numerous proposals have been advanced for improving the
11-20
quality of social control in the field of human experimentation. Some of these
proposals have suggested changes in the structure and/or functions of local
review committees, or Institutional Review Boards, as they are now officially
called. For example, some commentators have urged — and FDA regulations
require — that the membership of Institutional Review Boards be broadened to
54
include lay and non-scientif ic points of view. Additional responsibilities
have also been proposed for Institutional Review Boards , including the monitoring
of subject selection and the consent process and on-site evaluation of studies
55
in progress. Several new modes and mechanisms of social control are
intended to supplement the role of local Institutional Review Boards and
other existing review bodies. The National Commission for the Protection of
Human Subjects has been charged with multiple tasks, including the development
of national guidelines for the conduct of biomedical and behavioral research.
The Commission will also investigate whether subjects are adequately protected
in research contexts where current DHEW guidelines do not apply. Building
on the foundation laid by the Commission, soon-to-be-created Ethical Advisory
Boards will complement the efforts of Institutional Review Boards and other
review bodies by providing national review for particular categories of
research and by seeking to apply generally-accepted ethical standards to
57
complex individual cases.
As noted above, social control is a procedural rather than a substantive
issue. Therefore it is not in itself one of the necessary conditions for
ethically-acceptable human research. Rather, it is a means for ensuring, insofar
as possible, that research is conducted in accordance with basic ethical
principles. Thus, from an ethical perspective the primary criterion for
judging the adequacy of any particular social-control mechanism will be the
11-21
extent to which that mechanism promotes values like those identified in
earlier sections of this essay.
VII. Compensation of Injured Research Subjects
Even if a research project is exemplary in its design, risk-benefit ratio,
selection of subjects, and provision for informed consent, and even if the
project has been thoroughly reviewed and carefully monitored from inception
to implementation — it is possible that a participant in the project will
suffer injury. It is also possible, and indeed likely in a situation like
the one described, that the injury will have been due to no fault of the
investigator. Let us suppose, hypothethically , that such an unanticipated
and seemingly-unavoidable injury has occurred and, further, that it has been
suffered by a normal volunteer for whom the study in question was totally
nontherapeutic. What debt, if any, does society owe to the injured subject?
The major codes of ethics are totally silent on the subject of compensation
for research-related injuries. However, there is a significant and expanding
secondary literature which discusses the issue; most commentators have
58
advocated some form of compensation for injured subjects. In addition, an
inter-agency DHEW task force, the Secretary's Task Force on the Compensation
of Injured Research Subjects, is currently completing an intensive study
59
of this question.
In the initial section of this essay it was argued that, except in cases
of national emergency, no citizen should be required to take part in non-
therapeutic research. It would be possible to conclude that since society
does not mandate any subject's participation, neither does it bear responsibility
11-22
for the untoward effects of any individual's free decision, On the other hand,
however, the society requires that certain types of nontherapeutic research
be done, e.g., Phase I testing of most new drugs on normal volunteers. In
addition, the society supports many kinds of nontherapeutic research projects
which cannot be carried out unless substantial numbers of persons volunteer
to participate. Moreover, it can be argued that the freedom of the individual's
choice is a less significant consideration than the fact of the individual's
having been injured while contributing to the social good. In the context of
military service, society compensates both draftees and volunteers who suffer
service-connected disabilities. In an analogous way, the principle of justice
would seem to require compensation for casualties of the war on disease,
without regard to their volunteer status.
Even if arguments based on compensatory justice are rejected, the
probable implications of an expanded concept of informed consent may lead to
a similar conclusion. As noted above in Section V, some commentators have
urged that prospective subjects of nontherapeutic research be advised concerning
provisions for the repair of, or compensation for, research-related injuries.
If this type of information is in fact imparted to potential subjects, — as
it should be, in the author's view — then it is likely that increasing numbers
of subjects will request guarantees of restorative medical care or, in cases
where restoration is impossible, assurance of equitable compensation for
injuries.
Efforts to develop a mechanism for the equitable compensation of injured
research subjects will undoubtedly be confronted with numerous administrative
difficulties. The clearcut case of a no-fault injury suffered in the course
of totally-nontherapeutic research will be relatively straightforward in
11-23
terms of causation and presents a strong claim for compensation. Less clear
in both respects will be cases in which ill patient-subjects seem to suffer
injury while participating in mixed research (e.g. , research in which the
therapeutic/nontherapeutic ratio is 60/40) . In addition, for both simple
and complex cases an appropriate institutional framework will need to be
created and a calculus for determining appropriate levels of compensation will
need to be found.
Unlike the ethical principles developed in Sections II-V above, the
principle of equitable compensation for research-related injuries is a
contingent principle, which can only be applied in cases where actual injury
occurs. Thus, the requirement of equitable compensation can be categorized as
a subsidiary ethical principle and should be regarded as a necessary condition
for ethically-acceptable nontherapeutic research in cases where subjects of
that research suffer injury.
Conclusion
In the Summa theologiae of Thomas Aquinas there appears a question
entitled simply "On War." Thomas begins his discussion of that question by
asking, "Is it always wrong to wage war?" He answers in the negative but
adds that in order for a war to be just, i.e., morally right, "three things
are required": the authority of the prince, a just cause, and a right intention
on the part of those waging the war. Thomas apparently considered all three
of these criteria to be essential , for in discussing "right intention" he
noted:
It may happen that even if a war is declared by the
authority of the prince and even if the cause is just,
11-24
the war may nevertheless be rendered illicit
because of a wrong intention. 1
The formal structure of the present essay closely parallels Thomas'
question on war. Section I raised the issue whether nontherapeutic research
can be morally justified under any circumstances and concluded that such
justification is in principle possible. Sections II through V identified four
general requirements for ethically-acceptable nontherapeutic research: adequate
research design; a favorable risk-benefit ratio; equitable selection of subjects;
and reasonably free and adequately informed consent by the subjects. In Section
VI it was argued that social control is a procedural value which should assist
in the implementation of the four general requirements. And in Section VII
the subsidiary principle of equitable compensation for research-related injuries
was identified.
No attempt has been made to explore the relative importance of the four
general requirements discussed in Sections n through V. However, the language
of conditionality has been employed to assert the essential significance of
each. It is the thesis of this essay that the four general requirements —
adequate research design, favorable risk-benefit ratio, equitable subject
selection, and informed consent — are individually necessary and jointly
sufficient to provide moral justification for any proposed nontherapeutic
human research.
11-25
FOOTNOTES
This definition is adapted from a definition suggested by Robert
J. Levine in his essay for the Commission entitled "The Boundaries Between
Biomedical or Behavioral Research and the Accepted and Routine Practice
of Medicine," unpublished paper, 14 July 1975.
o
The definitions of therapeutic and nontherapeutic research closely
follow definitions proposed by Charles Fried ( Medical Experimentation :
Personal Integrity and Social Choice CNew York: American Elsevier, 19743,
pp. 25-26. The language of design is preferred to the language of intention
since it is more objective and less likely to become entangled with the
problem of self-deception.
Robert M. Veatch, "Ethical Principles in Medical Experimentation," in
Alice M. Rivlin and P. Michael Timpane, eds., Ethical and Legal Issues of
Social Experimentation (Washington: Brookings Institution, 1975) , pp. 24-26.
4 Fried, Medical Experimentation , pp. 47-60
Because of this apparent conflict in the physician-investigator's roles,
one physician has recently proposed that the two roles be rigorously
separated. See Howard M. Spiro, "Constraint and Consent — On Being a
Patient and a Subject," New England Journal of Medicine 293(22): 1134-1135,
27 November 1975.
Robert Q. Marston, "Medical Science, the Clinical Trial, and Society,"
address delivered at the University of Virginia, Charlottesville, Va. ,
November 10, 1972.
n
'William K. Frankena, Ethics (2nd ed.; Englewood Cliffs, N.J. :
Prentice-Hall, 1973), pp. 45-48. It is, of course, a further question whether
allowing harm to occur to a person is morally equivalent to causing harm to
that person.
g
Hans Jonas, "Philosophical Reflections on Human Experimentation," in
Paul A. Freund, ed. , Experimentation with Human Subjects (New York: George
Braziller, 1970) , pp. 6-15.
9
David D. Rutstein, "The Ethical Design of Human Experiments," in
Freund, ed. , Experimentation with Human Subjects , pp. 383-401; Fried,
Medical Experimentation , passim; Milton C. Weinstein, "Allocation of
Subjects in Medical Experiments," New England Journal of Medicine 291(24):
1278-1285, 12 December 1974; William W. May, "The Composition and Function
of Ethical Committees," Journal of Medical Ethics 1(1): 23-29, April 1975.
Rutstein, "The Ethical Design of Human Experiments," p. 384.
11-26
11 0ffice of the Secretary, DHEW, "Protection of Human Subjects:
Fetuses, Pregnant Women, and In Vitro Fertilization," Federal Register
40(154): 33545, 33547-33548, 8 August 1975.
12
William G. Cochran, "Experimental Design: I. The Design of Experiments,"
International Encyclopedia of the Social Sciences , Vol. 5, p. 246. See also
Ronald A. Fisher, The Design of Experiments (7th ed. ; New York: Hafner, 1960).
13
Weinstein, "Allocation of subjects in Medical Experiments," pp. 1279-1281;
Fried, Medical Experimentation , p. 159. See also Donald J. Campbell,
"Experimental Design: III. Quasi-Experimental Design," International
Encyclopedia of the Social Sciences , Vol. 14, pp. 259-263.
14
Lawrence W. Shaw and Thomas C. Chalmers, "Ethics in Cooperative Clinical
Trials," Annals of the New York Academy of Sciences 169 (Art. 2): 487-488,
21 January 1970.
Fried, Medical Experimentation, pp. 153-154.
Weinstein, "Allocation of Subjects in Medical Experiments," pp. 1281-
1284. See also P. Armitage, "Sequential Analysis," International Encyclopedia
of the Social Sciences , Vol. 14, pp. 187-192.
17 Weinstein, ibid . , p. 1284
18
On this issue see Fried, Medical Experimentation , p. 148
19
In this section the term "risk(s)" and "harm(s)" are used interchangeably,
and the concepts "risk (of harm) " and " (probability of) benefit" are considered
to be symmetrical.
20
Some of these problems have been more thoroughly explored in an
excellent paper prepared for the Commission by Robert J. Ievine ("The Role of
the Assessment of Risk-Benefit Criteria in the Determination of the
Appropriateness of Research Involving Human Subjects," unpublished paper,
27 October 1975) .
21
For a brief discussion of several problems inherent in cost-benefit
analysis see Fried, Medical Experimentation , pp. 81-87. See also
Nicholas Georgescu-Roegen, "Utility", International Encyclopedia of the
Social Sciences , Vol. 16, pp. 236-267.
22
The problem of collective choice has been vigorously debated by
welfare economists. For a recent survey of the various positions in this
debate which includes non-technical discussion and a comprehensive
bibliography see Amartya K. Sen, Collective Choice and Social Welfare
(San Francisco: Holden-Day, 1970) .
23
The Declaration of Helsinki employs the language of proportionality
in its third Basic Principle.
Jonas, "Philosophical Reflections on Human Experimentation," p. 18.
11-27
25 Ibid ., p. 20.
Ibid . , pp. 21-24.
27
Lawrence K. Altman, "Auto-Experimentation: An Unappreciated Tradition
in Medical Science," New England Journal of Medicine 286(7): 346-352,
17 February 1972; Fried, Medical Experimentation , pp. 167-168.
A.J. Benatt, "Cardiac Catheterization;" reprinted in Jay Katz , with the
assistance of Alexander Morgan Capron and Eleanor Swift Glass , Experimentation
with Human Beings (New York: Russell Sage Foundation, 1972) , p. 138.
29
Bernard Barber, et a^. , Research on Human Subjects; Problems of Social
Control in Medical Experimentation (New York: Russell Sage Foundation, 1973),
pp. 53-57; Katz, Experimentation with Human Beings , pp. 342-345, 633-634, 740-
746; Bradford H. Gray, Human Subjects in Medical Experimentation , pp. 241-243;
Fried, Medical Experimentation , pp. 165-171; Henry Foster, Jay Katz, and
Franz Ingelfinger, "The Poor," in Experiments and Research with Humans: Values
in Conflict (Washington: National Academy of Sciences, 1975), pp. 150-160.
Barber, e_t a_l. , Research on Human Subjects , pp. 54-56.
31
Fried, Medical Experimentation , pp. 169-170.
32
Paul Ramsey, The Patient as Person: Explorations in Medical Ethics
(New Haven, Yale University Press, 1970) , pp. 1-11; Jon R. Waltz and Thomas
W. Scheuneman, "Informed Consent to Therapy," reprinted in Katz, Experimentation
with Human Beings , pp. 579-581; Alexander M. Capron, "Informed Consent in
Catastrophic Disease Research and Treatment," University of Pennsylvania Law
Review 123(2) : 413-414, December 1974.
33
Ramsey, ibid . , p. 5.
34
An excerpt from the Schloendorff v. New York Hospital decision is
reprinted in Katz, Experimentation with Human Beings , p. 526. The sources
for the three 1972 decisions, respectively, are 502 P. 3d 1 (1972),
295 A. 2d 676 (1972), and 464 F.2d 772 (D.C. Cir.). For a thorough discussion
of these latter three cases see Capron, "Informed Consent in Catastrophic
Disease Research and Treatment," pp. 403-423.
35
Capron, ibid . , pp. 364-376.
36
The topics of proxy consent and substituted judgment are also highly
significant but are too complex to be analyzed satisfactorily in a survey essay.
For an excellent introduction to some of the ethical issues involved in the
analysis of these two topics, Richard A. McCormick, "Proxy Consent in the
Experimentation Situation," Perspectives in Biology and Medicine 18(1): 2-20,
Autumn 1974.
37
Franz J. Ingelfinger, "The Poor," in Experiments and Research with
Human Subjects , p. 169; Fried, Medical Experimentation, pp. 33-34.
11-28
38
Food and Drug Administration regulations stipulate that every subject
must be "provided with a fair explanation of pertinent information concerning
the investigational drug, and/or his possible use as a control " (Code of
Federal Regulations 21, 130.37 C 19713 ) . See also Veatch, "Ethical Principles
in Medical Experimentation," pp. 52-53.
39
Robert J. Levine, "The Nature and Definition of Informed Consent in
Various Research Settings," unpublished paper prepared for the National
Commission, December 1975, p. 16.
40
Ibid . ; Veatch, "Ethical Principles in Medical Experimentation," pp. 53-54.
41
Fried, Medical Experimentation , pp. 35-36. Fried's proposal illustrates
the fact that as a randomized comparison of two therapies approaches the
predetermined level of significance, the subjects assigned to the less desirable
therapy gradually approach the situation of normal volunteers in a totally-
nontherapeutic study.
42
Capron, "Informed Consent in Catastrophic Disease Research and Treatment,"
pp. 406-408.
43
502 P. 2d at 11; cited by Capron, ibid . , p. 406.
44
U.S. Code 21, 355 (1972).
45
Code of Federal Regulations 21, 130.37(b) and (f) (1971). See also Fried,
Medical Experimentation, pp. 39-41.
46
Levine, "The Nature and Definition of Informed Consent," pp. 30-31.
47
Levine argues for the possible moral justifiability of unseen-observer
research without specifying whether such research would be therapeutic or
nontherapeutic in its design ( ibid . , p. 78) .
48
Veatch, "Ethical Principles in Medical Experimentation," p. 49.
49 The words in brackets were substituted for "a committee of the organization"
in Technical Amendments to the May 30, 1974, guidelines which were published
in Federal Register 40(50): 11854-11558, 13 March 1975.
50 Gray, Human Subjects in Medical Experimentation , pp. 245-246; Barber,
et al . , Research on Human Subjects , pp. 173-174.
51 This judgment itself is a matter of considerable debate. See, for
example, Eliot Freidson, The Profession of Medicine (New York: Dodd, Mead,
1970) , chap. 7.
52 William J. Curran, "Government Regulation of the Use of Human Subjects
in Medical Research: The Approach of Two Federal Agencies," in Freund, ed. ,
Experimentation with Human Subjects, pp. 430-433.
53 Barber, Research on Human Subjects, p. 148.
11-29
54 Code of Federal Regulations 21, 130.3 (1971); first published in
Federal Register 32:8753, 20 June 1967. See also Veatch, "Ethical Principles
in Medical Experimentation," pp. 38-41.
Office of the Secretary, DHEW, "Protection of Human Subjects: Fetuses,
Pregnant Women, and In Vitro Fertilization," p. 33529 ( Code of Federal
Regulations 45, 46.205(2) (2) £197511).
Public Law 93-348, Section 202, U.S. Statutes at Large 88: 349-350.
Office of the Secretary, DHEW, "Protection of Human Subjects: Fetuses,
Pregnant Women, and In Vitro Fertilization," pp. 33526, 33529.
58
Fried, Medical Experimentation , pp. 26-28, 171-172; Clark C. Havighurst,
"Compensating Persons Injured in Human Experimentation," Science 169: 153-157,
10 July 1970; Irving Ladimer, "Clinical Research Insurance," Journal of
Chronic Disease 16: 1229-1235, 11 December 1963; Note, "Medical Experiment
Insurance," Columbia Law Review 70(5): 965-979, May 1970; Edwin Roth and Paul
Rothstein, "Non-Fault-Based Medical Injury Compensation Systems," in the Appendix
to the Report of the Secretary's Commission on Medical Malpractice (Washington
DHEW, 1973). pp. 450-493.
59 Personal communication from Dr. Charles R. McCarthy, Division of Legislative
Analysis, Office of the Director, NIH.
The question of compensating subjects injured in the course of mixed
research will be simplified if a general system of compensation for medical
accidents is established. On this point see Clark C. Havighurst and Laurence
R. Tancredi, "'Medical Adversity Insurance' — a No-Fault Approach to Medical
Malpractice and Quality Assurance," Milbank Memorial Fund Quarterly 51(2):
125-168, Spring 1973.
Summa theologiae 2-2, question 40, article 1.
11-30
• Lienjyty
N1H
l(pr Amazing Res
Research.
Amazing Help.
http://nihlibraty.nih.gov
10 Center Drive
Bethesda, MD 20892-1150
301-496-1080
DATE DUE
NIH LIBRARY
||f(| ! "f!|lf:"|,
4 01 20 "8964
NIH LIBRARY
3 J496 00113 2359
*' "*
U.S. Department of Health^Education, and Welfare
DHEW PublicationNo. (OS) 78-0013
Open Library
